Tokenizer: ../home/arg_mining/4epoch_complete/tokenizer Model: ../home/arg_mining/4epoch_complete/model/
	Train size: 80 Test size: 20


		-------------RUN 1-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.07933884297520662, 'partial': 0.3333333333333333}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 1.0, 'partial': 0.024390243902439025}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.14701378254211334, 'partial': 0.04545454545454545}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.0805921052631579, 'recall': 0.0805921052631579, 'f1': 0.0805921052631579, 'support': None}, 'macro_avg': {'precision': 0.08253443526170798, 'recall': 0.20487804878048782, 'f1': 0.038493665599331754, 'support': None}, 'weighted_avg': {'precision': 0.028741663041902277, 'recall': 0.0805921052631579, 'f1': 0.014671542640884544, 'support': None}}
Loss: tensor(1.8596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6012, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8345864661654135, 'agreement': 0.4074074074074074, 'direct_attack': 0.17592592592592593, 'undercutter_attack': 0.2826086956521739, 'partial': 0.0}, 'recall': {'support': 0.8387909319899244, 'agreement': 0.3548387096774194, 'direct_attack': 0.31666666666666665, 'undercutter_attack': 0.2708333333333333, 'partial': 0.0}, 'f1': {'support': 0.836683417085427, 'agreement': 0.3793103448275862, 'direct_attack': 0.2261904761904762, 'undercutter_attack': 0.2765957446808511, 'partial': 0.0}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6365131578947368, 'recall': 0.6365131578947368, 'f1': 0.6365131578947368, 'support': None}, 'macro_avg': {'precision': 0.3401056990301841, 'recall': 0.3562259283334687, 'f1': 0.3437559965568681, 'support': None}, 'weighted_avg': {'precision': 0.6261691764371519, 'recall': 0.6365131578947368, 'f1': 0.6291588524314708, 'support': None}}
Loss: tensor(2.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1951, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8872549019607843, 'agreement': 0.3235294117647059, 'direct_attack': 0.2571428571428571, 'undercutter_attack': 0.35555555555555557, 'partial': 0.42857142857142855}, 'recall': {'support': 0.9095477386934674, 'agreement': 0.3548387096774194, 'direct_attack': 0.15, 'undercutter_attack': 0.6666666666666666, 'partial': 0.075}, 'f1': {'support': 0.8982630272952854, 'agreement': 0.3384615384615384, 'direct_attack': 0.1894736842105263, 'undercutter_attack': 0.46376811594202905, 'partial': 0.1276595744680851}, 'support': {'support': 398, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 40}, 'micro_avg': {'precision': 0.7039473684210527, 'recall': 0.7039473684210527, 'f1': 0.7039473684210527, 'support': None}, 'macro_avg': {'precision': 0.4504108309990663, 'recall': 0.4312106230075107, 'f1': 0.4035251880754928, 'support': None}, 'weighted_avg': {'precision': 0.6954348186643079, 'recall': 0.7039473684210527, 'f1': 0.6862318648761698, 'support': None}}
Loss: tensor(0.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9051282051282051, 'agreement': 0.3620689655172414, 'direct_attack': 0.16, 'undercutter_attack': 0.17647058823529413, 'partial': 0.1864406779661017}, 'recall': {'support': 0.889168765743073, 'agreement': 0.3387096774193548, 'direct_attack': 0.13333333333333333, 'undercutter_attack': 0.1875, 'partial': 0.2682926829268293}, 'f1': {'support': 0.8970775095298602, 'agreement': 0.35, 'direct_attack': 0.14545454545454545, 'undercutter_attack': 0.1818181818181818, 'partial': 0.22}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6611842105263158, 'recall': 0.6611842105263158, 'f1': 0.6611842105263158, 'support': None}, 'macro_avg': {'precision': 0.35802168736936846, 'recall': 0.3634008918845181, 'f1': 0.3588700473605175, 'support': None}, 'weighted_avg': {'precision': 0.6702283377136031, 'recall': 0.6611842105263158, 'f1': 0.6649906525294407, 'support': None}}
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9122807017543859, 'agreement': 0.358974358974359, 'direct_attack': 0.2916666666666667, 'undercutter_attack': 0.4057971014492754, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9168765743073047, 'agreement': 0.45161290322580644, 'direct_attack': 0.23333333333333334, 'undercutter_attack': 0.5833333333333334, 'partial': 0.04878048780487805}, 'f1': {'support': 0.9145728643216079, 'agreement': 0.4, 'direct_attack': 0.2592592592592593, 'undercutter_attack': 0.47863247863247865, 'partial': 0.07272727272727274}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.7171052631578947, 'recall': 0.7171052631578947, 'f1': 0.7171052631578947, 'support': None}, 'macro_avg': {'precision': 0.42231519434036596, 'recall': 0.4467873264009312, 'f1': 0.42503837498812375, 'support': None}, 'weighted_avg': {'precision': 0.7027421917427789, 'recall': 0.7171052631578947, 'f1': 0.7062453286963998, 'support': None}}
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8976982097186701, 'agreement': 0.3275862068965517, 'direct_attack': 0.2601626016260163, 'undercutter_attack': 0.10526315789473684, 'partial': 0.17647058823529413}, 'recall': {'support': 0.8841309823677582, 'agreement': 0.3064516129032258, 'direct_attack': 0.5333333333333333, 'undercutter_attack': 0.041666666666666664, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8908629441624367, 'agreement': 0.31666666666666665, 'direct_attack': 0.34972677595628415, 'undercutter_attack': 0.05970149253731344, 'partial': 0.10344827586206896}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6694078947368421, 'recall': 0.6694078947368421, 'f1': 0.6694078947368421, 'support': None}, 'macro_avg': {'precision': 0.3534361528742538, 'recall': 0.36775066539566026, 'f1': 0.344081231036954, 'support': None}, 'weighted_avg': {'precision': 0.6654510129606145, 'recall': 0.6694078947368421, 'f1': 0.660191742887062, 'support': None}}
Loss: tensor(1.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9020618556701031, 'agreement': 0.3013698630136986, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.31932773109243695, 'partial': 0.6}, 'recall': {'support': 0.8816120906801007, 'agreement': 0.3548387096774194, 'direct_attack': 0.1, 'undercutter_attack': 0.7916666666666666, 'partial': 0.14634146341463414}, 'f1': {'support': 0.89171974522293, 'agreement': 0.32592592592592595, 'direct_attack': 0.15384615384615383, 'undercutter_attack': 0.4550898203592814, 'partial': 0.23529411764705882}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.694078947368421, 'recall': 0.694078947368421, 'f1': 0.694078947368421, 'support': None}, 'macro_avg': {'precision': 0.4912185566219144, 'recall': 0.45489178608776426, 'f1': 0.41237515260026997, 'support': None}, 'weighted_avg': {'precision': 0.7183079264807849, 'recall': 0.694078947368421, 'f1': 0.6824708646257479, 'support': None}}
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.889168765743073, 'agreement': 0.3, 'direct_attack': 0.2727272727272727, 'undercutter_attack': 0.32653061224489793, 'partial': 0.35}, 'recall': {'support': 0.889168765743073, 'agreement': 0.2903225806451613, 'direct_attack': 0.15, 'undercutter_attack': 0.6666666666666666, 'partial': 0.17073170731707318}, 'f1': {'support': 0.889168765743073, 'agreement': 0.29508196721311475, 'direct_attack': 0.19354838709677416, 'undercutter_attack': 0.43835616438356156, 'partial': 0.22950819672131145}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6891447368421053, 'recall': 0.6891447368421053, 'f1': 0.6891447368421053, 'support': None}, 'macro_avg': {'precision': 0.42768533014304866, 'recall': 0.4333779440743949, 'f1': 0.409132696231567, 'support': None}, 'weighted_avg': {'precision': 0.6874787923542623, 'recall': 0.6891447368421053, 'f1': 0.6798666400477044, 'support': None}}
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8886138613861386, 'agreement': 0.27692307692307694, 'direct_attack': 0.34234234234234234, 'undercutter_attack': 0.35, 'partial': 0.25}, 'recall': {'support': 0.9042821158690176, 'agreement': 0.2903225806451613, 'direct_attack': 0.6333333333333333, 'undercutter_attack': 0.14583333333333334, 'partial': 0.04878048780487805}, 'f1': {'support': 0.8963795255930087, 'agreement': 0.28346456692913385, 'direct_attack': 0.4444444444444444, 'undercutter_attack': 0.20588235294117646, 'partial': 0.0816326530612245}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6973684210526315, 'recall': 0.6973684210526315, 'f1': 0.6973684210526315, 'support': None}, 'macro_avg': {'precision': 0.4215758561303116, 'recall': 0.40451037019714475, 'f1': 0.38236070859379756, 'support': None}, 'weighted_avg': {'precision': 0.6867425563816912, 'recall': 0.6973684210526315, 'f1': 0.6798247256470132, 'support': None}}
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9020618556701031, 'agreement': 0.2962962962962963, 'direct_attack': 0.39603960396039606, 'undercutter_attack': 0.4074074074074074, 'partial': 0.45454545454545453}, 'recall': {'support': 0.8816120906801007, 'agreement': 0.3870967741935484, 'direct_attack': 0.6666666666666666, 'undercutter_attack': 0.22916666666666666, 'partial': 0.12195121951219512}, 'f1': {'support': 0.89171974522293, 'agreement': 0.3356643356643356, 'direct_attack': 0.49689440993788814, 'undercutter_attack': 0.2933333333333333, 'partial': 0.19230769230769232}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.7072368421052632, 'recall': 0.7072368421052632, 'f1': 0.7072368421052632, 'support': None}, 'macro_avg': {'precision': 0.49127012357593147, 'recall': 0.4572986835438355, 'f1': 0.44198390329323595, 'support': None}, 'weighted_avg': {'precision': 0.7211237212186582, 'recall': 0.7072368421052632, 'f1': 0.701648367838126, 'support': None}}
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9178082191780822, 'agreement': 0.24475524475524477, 'direct_attack': 0.3050847457627119, 'undercutter_attack': 0.3235294117647059, 'partial': 0.2857142857142857}, 'recall': {'support': 0.8417085427135679, 'agreement': 0.5645161290322581, 'direct_attack': 0.3050847457627119, 'undercutter_attack': 0.22916666666666666, 'partial': 0.04878048780487805}, 'f1': {'support': 0.8781127129750983, 'agreement': 0.34146341463414637, 'direct_attack': 0.3050847457627119, 'undercutter_attack': 0.26829268292682923, 'partial': 0.08333333333333334}, 'support': {'support': 398, 'agreement': 62, 'direct_attack': 59, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6595394736842105, 'recall': 0.6595394736842105, 'f1': 0.6595394736842105, 'support': None}, 'macro_avg': {'precision': 0.4153783814350061, 'recall': 0.39785131439601656, 'f1': 0.37525737792642383, 'support': None}, 'weighted_avg': {'precision': 0.7001746609978511, 'recall': 0.6595394736842105, 'f1': 0.666043267958159, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9273743016759777, 'agreement': 0.2767857142857143, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.3482142857142857, 'partial': 0.3684210526315789}, 'recall': {'support': 0.8341708542713567, 'agreement': 0.5081967213114754, 'direct_attack': 0.03333333333333333, 'undercutter_attack': 0.8125, 'partial': 0.17073170731707318}, 'f1': {'support': 0.8783068783068783, 'agreement': 0.3583815028901734, 'direct_attack': 0.05970149253731343, 'undercutter_attack': 0.48750000000000004, 'partial': 0.23333333333333336}, 'support': {'support': 398, 'agreement': 61, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6759868421052632, 'recall': 0.6759868421052632, 'f1': 0.6759868421052632, 'support': None}, 'macro_avg': {'precision': 0.44130192800436835, 'recall': 0.47178652324664777, 'f1': 0.4034446414135397, 'support': None}, 'weighted_avg': {'precision': 0.7153639912064232, 'recall': 0.6759868421052632, 'f1': 0.6710134300351046, 'support': None}}
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9093264248704663, 'agreement': 0.30120481927710846, 'direct_attack': 0.3368421052631579, 'undercutter_attack': 0.3888888888888889, 'partial': 0.375}, 'recall': {'support': 0.8841309823677582, 'agreement': 0.4032258064516129, 'direct_attack': 0.5333333333333333, 'undercutter_attack': 0.2916666666666667, 'partial': 0.07317073170731707}, 'f1': {'support': 0.896551724137931, 'agreement': 0.3448275862068965, 'direct_attack': 0.41290322580645156, 'undercutter_attack': 0.3333333333333333, 'partial': 0.12244897959183672}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6990131578947368, 'recall': 0.6990131578947368, 'f1': 0.6990131578947368, 'support': None}, 'macro_avg': {'precision': 0.4622524476599243, 'recall': 0.4371055041053376, 'f1': 0.4220129698152898, 'support': None}, 'weighted_avg': {'precision': 0.7136998066631777, 'recall': 0.6990131578947368, 'f1': 0.6958962936500633, 'support': None}}
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9128065395095368, 'agreement': 0.2857142857142857, 'direct_attack': 0.3717948717948718, 'undercutter_attack': 0.42105263157894735, 'partial': 0.25}, 'recall': {'support': 0.8438287153652393, 'agreement': 0.45161290322580644, 'direct_attack': 0.48333333333333334, 'undercutter_attack': 0.5, 'partial': 0.04878048780487805}, 'f1': {'support': 0.8769633507853404, 'agreement': 0.35000000000000003, 'direct_attack': 0.42028985507246375, 'undercutter_attack': 0.45714285714285713, 'partial': 0.0816326530612245}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6875, 'recall': 0.6875, 'f1': 0.6875, 'support': None}, 'macro_avg': {'precision': 0.4482736657195283, 'recall': 0.46551108794585144, 'f1': 0.4372057432123772, 'support': None}, 'weighted_avg': {'precision': 0.7119518100708119, 'recall': 0.6875, 'f1': 0.6913842721784461, 'support': None}}
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.904639175257732, 'agreement': 0.22033898305084745, 'direct_attack': 0.32558139534883723, 'undercutter_attack': 0.2222222222222222, 'partial': 0.24390243902439024}, 'recall': {'support': 0.8819095477386935, 'agreement': 0.4262295081967213, 'direct_attack': 0.23333333333333334, 'undercutter_attack': 0.08333333333333333, 'partial': 0.24390243902439024}, 'f1': {'support': 0.8931297709923665, 'agreement': 0.29050279329608936, 'direct_attack': 0.2718446601941748, 'undercutter_attack': 0.12121212121212122, 'partial': 0.24390243902439024}, 'support': {'support': 398, 'agreement': 61, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6661184210526315, 'recall': 0.6661184210526315, 'f1': 0.6661184210526315, 'support': None}, 'macro_avg': {'precision': 0.38333684298080584, 'recall': 0.37374163232529434, 'f1': 0.36411835694382844, 'support': None}, 'weighted_avg': {'precision': 0.6804089146484802, 'recall': 0.6661184210526315, 'f1': 0.6666368103221308, 'support': None}}
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8875, 'agreement': 0.2835820895522388, 'direct_attack': 0.32, 'undercutter_attack': 0.3333333333333333, 'partial': 0.42105263157894735}, 'recall': {'support': 0.8942065491183879, 'agreement': 0.3064516129032258, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.5, 'partial': 0.1951219512195122}, 'f1': {'support': 0.890840652446675, 'agreement': 0.2945736434108527, 'direct_attack': 0.2909090909090909, 'undercutter_attack': 0.4, 'partial': 0.26666666666666666}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.694078947368421, 'recall': 0.694078947368421, 'f1': 0.694078947368421, 'support': None}, 'macro_avg': {'precision': 0.4490936108929039, 'recall': 0.4324893559815585, 'f1': 0.42859801068665704, 'support': None}, 'weighted_avg': {'precision': 0.6947084661956836, 'recall': 0.694078947368421, 'f1': 0.6899920784550684, 'support': None}}
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8807785888077859, 'agreement': 0.3333333333333333, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.4117647058823529, 'partial': 0.42857142857142855}, 'recall': {'support': 0.9118387909319899, 'agreement': 0.3387096774193548, 'direct_attack': 0.4, 'undercutter_attack': 0.4375, 'partial': 0.07317073170731707}, 'f1': {'support': 0.896039603960396, 'agreement': 0.33599999999999997, 'direct_attack': 0.35294117647058826, 'undercutter_attack': 0.42424242424242425, 'partial': 0.125}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.7088815789473685, 'recall': 0.7088815789473685, 'f1': 0.7088815789473685, 'support': None}, 'macro_avg': {'precision': 0.47404750605582213, 'recall': 0.43224384001173244, 'f1': 0.42684464093468166, 'support': None}, 'weighted_avg': {'precision': 0.7016764297667629, 'recall': 0.7088815789473685, 'f1': 0.6960934699410344, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8989637305699482, 'agreement': 0.27184466019417475, 'direct_attack': 0.25, 'undercutter_attack': 0.35185185185185186, 'partial': 0.3333333333333333}, 'recall': {'support': 0.8740554156171285, 'agreement': 0.45161290322580644, 'direct_attack': 0.03333333333333333, 'undercutter_attack': 0.7916666666666666, 'partial': 0.024390243902439025}, 'f1': {'support': 0.8863346104725415, 'agreement': 0.3393939393939394, 'direct_attack': 0.058823529411764705, 'undercutter_attack': 0.48717948717948717, 'partial': 0.04545454545454545}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6842105263157895, 'recall': 0.6842105263157895, 'f1': 0.6842105263157895, 'support': None}, 'macro_avg': {'precision': 0.4211987151898616, 'recall': 0.43501171254907484, 'f1': 0.3634372223824557, 'support': None}, 'weighted_avg': {'precision': 0.6896357327695128, 'recall': 0.6842105263157895, 'f1': 0.6606824475542448, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9330e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9383259911894273, 'agreement': 0.233502538071066, 'direct_attack': 0.23529411764705882, 'undercutter_attack': 0.2087912087912088, 'partial': 0.125}, 'recall': {'support': 0.535175879396985, 'agreement': 0.7419354838709677, 'direct_attack': 0.3389830508474576, 'undercutter_attack': 0.3958333333333333, 'partial': 0.024390243902439025}, 'f1': {'support': 0.6816, 'agreement': 0.3552123552123552, 'direct_attack': 0.27777777777777773, 'undercutter_attack': 0.2733812949640288, 'partial': 0.04081632653061225}, 'support': {'support': 398, 'agreement': 62, 'direct_attack': 59, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.4917763157894737, 'recall': 0.4917763157894737, 'f1': 0.4917763157894737, 'support': None}, 'macro_avg': {'precision': 0.34818277113975216, 'recall': 0.40726359827023656, 'f1': 0.3257575508969548, 'support': None}, 'weighted_avg': {'precision': 0.68578985660683, 'recall': 0.4917763157894737, 'f1': 0.533691819832374, 'support': None}}
Loss: tensor(0.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(11.8597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9083769633507853, 'agreement': 0.2857142857142857, 'direct_attack': 0.47368421052631576, 'undercutter_attack': 0.4, 'partial': 0.5}, 'recall': {'support': 0.8740554156171285, 'agreement': 0.5161290322580645, 'direct_attack': 0.3, 'undercutter_attack': 0.5, 'partial': 0.1951219512195122}, 'f1': {'support': 0.8908857509627728, 'agreement': 0.367816091954023, 'direct_attack': 0.3673469387755102, 'undercutter_attack': 0.4444444444444445, 'partial': 0.28070175438596495}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.7055921052631579, 'recall': 0.7055921052631579, 'f1': 0.7055921052631579, 'support': None}, 'macro_avg': {'precision': 0.5135550919182774, 'recall': 0.47706127981894114, 'f1': 0.4702389961045431, 'support': None}, 'weighted_avg': {'precision': 0.7343108434146817, 'recall': 0.7055921052631579, 'f1': 0.7094887539852939, 'support': None}}
Loss: tensor(7.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9266862170087976, 'agreement': 0.21176470588235294, 'direct_attack': 0.3448275862068966, 'undercutter_attack': 0.23333333333333334, 'partial': 0.2222222222222222}, 'recall': {'support': 0.7959697732997482, 'agreement': 0.5806451612903226, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.14583333333333334, 'partial': 0.04878048780487805}, 'f1': {'support': 0.8563685636856367, 'agreement': 0.3103448275862069, 'direct_attack': 0.33898305084745767, 'undercutter_attack': 0.17948717948717952, 'partial': 0.08}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6266447368421053, 'recall': 0.6266447368421053, 'f1': 0.6266447368421053, 'support': None}, 'macro_avg': {'precision': 0.3877668129307206, 'recall': 0.38091241781232305, 'f1': 0.3530367243212962, 'support': None}, 'weighted_avg': {'precision': 0.694119418093295, 'recall': 0.6266447368421053, 'f1': 0.6438389255917346, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9273255813953488, 'agreement': 0.24087591240875914, 'direct_attack': 0.3287671232876712, 'undercutter_attack': 0.3111111111111111, 'partial': 0.3333333333333333}, 'recall': {'support': 0.8015075376884422, 'agreement': 0.5409836065573771, 'direct_attack': 0.4, 'undercutter_attack': 0.2916666666666667, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8598382749326144, 'agreement': 0.33333333333333337, 'direct_attack': 0.3609022556390977, 'undercutter_attack': 0.30107526881720426, 'partial': 0.12}, 'support': {'support': 398, 'agreement': 61, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6463815789473685, 'recall': 0.6463815789473685, 'f1': 0.6463815789473685, 'support': None}, 'macro_avg': {'precision': 0.42828261230724474, 'recall': 0.42146570852396054, 'f1': 0.39502982654444996, 'support': None}, 'weighted_avg': {'precision': 0.7106826306735912, 'recall': 0.6463815789473685, 'f1': 0.6637742022994827, 'support': None}}
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9177718832891246, 'agreement': 0.29591836734693877, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.38666666666666666, 'partial': 0.3333333333333333}, 'recall': {'support': 0.8693467336683417, 'agreement': 0.46774193548387094, 'direct_attack': 0.23333333333333334, 'undercutter_attack': 0.6041666666666666, 'partial': 0.075}, 'f1': {'support': 0.8929032258064515, 'agreement': 0.36249999999999993, 'direct_attack': 0.25688073394495414, 'undercutter_attack': 0.4715447154471545, 'partial': 0.12244897959183673}, 'support': {'support': 398, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 40}, 'micro_avg': {'precision': 0.6924342105263158, 'recall': 0.6924342105263158, 'f1': 0.6924342105263158, 'support': None}, 'macro_avg': {'precision': 0.4438809072700699, 'recall': 0.4499177338304426, 'f1': 0.4212555309580794, 'support': None}, 'weighted_avg': {'precision': 0.7116058203960071, 'recall': 0.6924342105263158, 'f1': 0.6920977523565819, 'support': None}}
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9173333333333333, 'agreement': 0.29292929292929293, 'direct_attack': 0.2830188679245283, 'undercutter_attack': 0.39436619718309857, 'partial': 0.3}, 'recall': {'support': 0.8664987405541562, 'agreement': 0.46774193548387094, 'direct_attack': 0.25, 'undercutter_attack': 0.5833333333333334, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8911917098445596, 'agreement': 0.36024844720496896, 'direct_attack': 0.26548672566371684, 'undercutter_attack': 0.4705882352941176, 'partial': 0.11764705882352941}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6891447368421053, 'recall': 0.6891447368421053, 'f1': 0.6891447368421053, 'support': None}, 'macro_avg': {'precision': 0.4375295382740506, 'recall': 0.4481489482157356, 'f1': 0.4210324353661784, 'support': None}, 'weighted_avg': {'precision': 0.7081474655184374, 'recall': 0.6891447368421053, 'f1': 0.6899333565472098, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9171122994652406, 'agreement': 0.29, 'direct_attack': 0.2830188679245283, 'undercutter_attack': 0.39436619718309857, 'partial': 0.3}, 'recall': {'support': 0.8639798488664987, 'agreement': 0.46774193548387094, 'direct_attack': 0.25, 'undercutter_attack': 0.5833333333333334, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8897535667963683, 'agreement': 0.35802469135802467, 'direct_attack': 0.26548672566371684, 'undercutter_attack': 0.4705882352941176, 'partial': 0.11764705882352941}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6875, 'recall': 0.6875, 'f1': 0.6875, 'support': None}, 'macro_avg': {'precision': 0.4368994729145735, 'recall': 0.447645169878204, 'f1': 0.42030005558715133, 'support': None}, 'weighted_avg': {'precision': 0.7077044283354622, 'recall': 0.6875, 'f1': 0.6887675413290478, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4772e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9173333333333333, 'agreement': 0.29, 'direct_attack': 0.2830188679245283, 'undercutter_attack': 0.38571428571428573, 'partial': 0.3}, 'recall': {'support': 0.864321608040201, 'agreement': 0.46774193548387094, 'direct_attack': 0.25, 'undercutter_attack': 0.574468085106383, 'partial': 0.07317073170731707}, 'f1': {'support': 0.890038809831824, 'agreement': 0.35802469135802467, 'direct_attack': 0.26548672566371684, 'undercutter_attack': 0.46153846153846156, 'partial': 0.11764705882352941}, 'support': {'support': 398, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 47, 'partial': 41}, 'micro_avg': {'precision': 0.6875, 'recall': 0.6875, 'f1': 0.6875, 'support': None}, 'macro_avg': {'precision': 0.43521329739442943, 'recall': 0.44594047206755444, 'f1': 0.4185471494431113, 'support': None}, 'weighted_avg': {'precision': 0.7080400825176149, 'recall': 0.6875, 'f1': 0.688944108258485, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4772e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7808e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9171122994652406, 'agreement': 0.29292929292929293, 'direct_attack': 0.2830188679245283, 'undercutter_attack': 0.4027777777777778, 'partial': 0.3}, 'recall': {'support': 0.8639798488664987, 'agreement': 0.46774193548387094, 'direct_attack': 0.25, 'undercutter_attack': 0.6041666666666666, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8897535667963683, 'agreement': 0.36024844720496896, 'direct_attack': 0.26548672566371684, 'undercutter_attack': 0.48333333333333334, 'partial': 0.11764705882352941}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6891447368421053, 'recall': 0.6891447368421053, 'f1': 0.6891447368421053, 'support': None}, 'macro_avg': {'precision': 0.4391676476193679, 'recall': 0.4518118365448707, 'f1': 0.4232938263643834, 'support': None}, 'weighted_avg': {'precision': 0.7086672112798055, 'recall': 0.6891447368421053, 'f1': 0.6900004978560098, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9330e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6613e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9171122994652406, 'agreement': 0.29591836734693877, 'direct_attack': 0.2962962962962963, 'undercutter_attack': 0.4027777777777778, 'partial': 0.3}, 'recall': {'support': 0.8639798488664987, 'agreement': 0.46774193548387094, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.6041666666666666, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8897535667963683, 'agreement': 0.36249999999999993, 'direct_attack': 0.28070175438596495, 'undercutter_attack': 0.48333333333333334, 'partial': 0.11764705882352941}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6907894736842105, 'recall': 0.6907894736842105, 'f1': 0.6907894736842105, 'support': None}, 'macro_avg': {'precision': 0.44242094817725064, 'recall': 0.455145169878204, 'f1': 0.4267871426678392, 'support': None}, 'weighted_avg': {'precision': 0.7102822907472399, 'recall': 0.6907894736842105, 'f1': 0.691731580087304, 'support': None}}
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6500e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3569e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3569e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6605e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6621e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7392e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9171122994652406, 'agreement': 0.29591836734693877, 'direct_attack': 0.2962962962962963, 'undercutter_attack': 0.39436619718309857, 'partial': 0.2727272727272727}, 'recall': {'support': 0.8639798488664987, 'agreement': 0.46774193548387094, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.5833333333333334, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8897535667963683, 'agreement': 0.36249999999999993, 'direct_attack': 0.28070175438596495, 'undercutter_attack': 0.4705882352941176, 'partial': 0.11538461538461536}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6891447368421053, 'recall': 0.6891447368421053, 'f1': 0.6891447368421053, 'support': None}, 'macro_avg': {'precision': 0.4352840866037694, 'recall': 0.4509785032115373, 'f1': 0.4237856343722132, 'support': None}, 'weighted_avg': {'precision': 0.7077791037624925, 'recall': 0.6891447368421053, 'f1': 0.690572822049676, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9019e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0398e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9665e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-07, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9148936170212766, 'agreement': 0.29591836734693877, 'direct_attack': 0.3090909090909091, 'undercutter_attack': 0.391304347826087, 'partial': 0.3}, 'recall': {'support': 0.864321608040201, 'agreement': 0.46774193548387094, 'direct_attack': 0.288135593220339, 'undercutter_attack': 0.5625, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8888888888888888, 'agreement': 0.36249999999999993, 'direct_attack': 0.2982456140350877, 'undercutter_attack': 0.46153846153846156, 'partial': 0.11764705882352941}, 'support': {'support': 398, 'agreement': 62, 'direct_attack': 59, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6907894736842105, 'recall': 0.6907894736842105, 'f1': 0.6907894736842105, 'support': None}, 'macro_avg': {'precision': 0.44224144825704226, 'recall': 0.4511739736903456, 'f1': 0.4257640046571935, 'support': None}, 'weighted_avg': {'precision': 0.7101867938848587, 'recall': 0.6907894736842105, 'f1': 0.6921490864662151, 'support': None}}
Loss: tensor(6.5911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6613e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7816e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9146666666666666, 'agreement': 0.29591836734693877, 'direct_attack': 0.32727272727272727, 'undercutter_attack': 0.391304347826087, 'partial': 0.2727272727272727}, 'recall': {'support': 0.8639798488664987, 'agreement': 0.46774193548387094, 'direct_attack': 0.3, 'undercutter_attack': 0.5625, 'partial': 0.07317073170731707}, 'f1': {'support': 0.8886010362694301, 'agreement': 0.36249999999999993, 'direct_attack': 0.3130434782608696, 'undercutter_attack': 0.46153846153846156, 'partial': 0.11538461538461536}, 'support': {'support': 397, 'agreement': 62, 'direct_attack': 60, 'undercutter_attack': 48, 'partial': 41}, 'micro_avg': {'precision': 0.6907894736842105, 'recall': 0.6907894736842105, 'f1': 0.6907894736842105, 'support': None}, 'macro_avg': {'precision': 0.44037787636793835, 'recall': 0.4534785032115374, 'f1': 0.42821351829067533, 'support': None}, 'weighted_avg': {'precision': 0.7089973617697547, 'recall': 0.6907894736842105, 'f1': 0.6922974267750516, 'support': None}}


		-------------RUN 2-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.0, 'agreement': 0.145610278372591, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.05263157894736842}, 'recall': {'support': 0.0, 'agreement': 0.8, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.09615384615384616}, 'f1': {'support': 0.0, 'agreement': 0.24637681159420288, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.06802721088435375}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.1280701754385965, 'recall': 0.1280701754385965, 'f1': 0.1280701754385965, 'support': None}, 'macro_avg': {'precision': 0.03964837146399188, 'recall': 0.17923076923076925, 'f1': 0.06288080449571133, 'support': None}, 'weighted_avg': {'precision': 0.02651529081918139, 'recall': 0.1280701754385965, 'f1': 0.042946392897357265, 'support': None}}
Loss: tensor(0.9807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6052, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8442367601246106, 'agreement': 0.5454545454545454, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.5333333333333333, 'partial': 0.3}, 'recall': {'support': 0.9377162629757786, 'agreement': 0.35294117647058826, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.5714285714285714, 'partial': 0.17307692307692307}, 'f1': {'support': 0.8885245901639344, 'agreement': 0.42857142857142855, 'direct_attack': 0.29850746268656714, 'undercutter_attack': 0.5517241379310344, 'partial': 0.21951219512195122}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6631578947368421, 'recall': 0.6631578947368421, 'f1': 0.6631578947368421, 'support': None}, 'macro_avg': {'precision': 0.4986589818365519, 'recall': 0.4736992534570389, 'f1': 0.4773679628949831, 'support': None}, 'weighted_avg': {'precision': 0.6437969758874825, 'recall': 0.6631578947368421, 'f1': 0.6471615569819137, 'support': None}}
Loss: tensor(1.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(11.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0975, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.888030888030888, 'agreement': 0.3333333333333333, 'direct_attack': 0.29577464788732394, 'undercutter_attack': 0.6666666666666666, 'partial': 0.22580645161290322}, 'recall': {'support': 0.7958477508650519, 'agreement': 0.5176470588235295, 'direct_attack': 0.7, 'undercutter_attack': 0.047619047619047616, 'partial': 0.1346153846153846}, 'f1': {'support': 0.8394160583941606, 'agreement': 0.40552995391705066, 'direct_attack': 0.41584158415841577, 'undercutter_attack': 0.08888888888888889, 'partial': 0.1686746987951807}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.5736842105263158, 'recall': 0.5736842105263158, 'f1': 0.5736842105263158, 'support': None}, 'macro_avg': {'precision': 0.481922397506223, 'recall': 0.43914584838460274, 'f1': 0.3836702368307393, 'support': None}, 'weighted_avg': {'precision': 0.6499345163708252, 'recall': 0.5736842105263158, 'f1': 0.5583325140568117, 'support': None}}
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8284883720930233, 'agreement': 0.59375, 'direct_attack': 0.2919254658385093, 'undercutter_attack': 0.5161290322580645, 'partial': 1.0}, 'recall': {'support': 0.986159169550173, 'agreement': 0.2235294117647059, 'direct_attack': 0.7833333333333333, 'undercutter_attack': 0.19047619047619047, 'partial': 0.038461538461538464}, 'f1': {'support': 0.9004739336492891, 'agreement': 0.3247863247863248, 'direct_attack': 0.42533936651583704, 'undercutter_attack': 0.27826086956521734, 'partial': 0.07407407407407407}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6473684210526316, 'recall': 0.6473684210526316, 'f1': 0.6473684210526316, 'support': None}, 'macro_avg': {'precision': 0.6460585740379194, 'recall': 0.4443919287171882, 'f1': 0.4005869137181485, 'support': None}, 'weighted_avg': {'precision': 0.7066179933243364, 'recall': 0.6473684210526316, 'f1': 0.59752619529432, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8705501618122977, 'agreement': 0.49333333333333335, 'direct_attack': 0.38461538461538464, 'undercutter_attack': 0.4215686274509804, 'partial': 0.1875}, 'recall': {'support': 0.9307958477508651, 'agreement': 0.43529411764705883, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.5119047619047619, 'partial': 0.11538461538461539}, 'f1': {'support': 0.899665551839465, 'agreement': 0.4625, 'direct_attack': 0.3571428571428571, 'undercutter_attack': 0.46236559139784944, 'partial': 0.14285714285714285}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6578947368421053, 'recall': 0.6578947368421053, 'f1': 0.6578947368421053, 'support': None}, 'macro_avg': {'precision': 0.4715135014423992, 'recall': 0.46534253520412683, 'f1': 0.4649062286474629, 'support': None}, 'weighted_avg': {'precision': 0.6346684524208647, 'recall': 0.6578947368421053, 'f1': 0.6438801702038028, 'support': None}}
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8299120234604106, 'agreement': 0.5454545454545454, 'direct_attack': 0.3541666666666667, 'undercutter_attack': 0.48333333333333334, 'partial': 0.23529411764705882}, 'recall': {'support': 0.9792387543252595, 'agreement': 0.1411764705882353, 'direct_attack': 0.5666666666666667, 'undercutter_attack': 0.34523809523809523, 'partial': 0.23076923076923078}, 'f1': {'support': 0.8984126984126984, 'agreement': 0.22429906542056074, 'direct_attack': 0.4358974358974359, 'undercutter_attack': 0.4027777777777778, 'partial': 0.23300970873786409}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6491228070175439, 'recall': 0.6491228070175439, 'f1': 0.6491228070175439, 'support': None}, 'macro_avg': {'precision': 0.48963213731240296, 'recall': 0.4526178435174975, 'f1': 0.43887933724926737, 'support': None}, 'weighted_avg': {'precision': 0.6320938688795475, 'recall': 0.6491228070175439, 'f1': 0.6154567977957296, 'support': None}}
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.835820895522388, 'agreement': 0.4533333333333333, 'direct_attack': 0.30833333333333335, 'undercutter_attack': 0.5555555555555556, 'partial': 1.0}, 'recall': {'support': 0.9688581314878892, 'agreement': 0.4, 'direct_attack': 0.6166666666666667, 'undercutter_attack': 0.23809523809523808, 'partial': 0.07692307692307693}, 'f1': {'support': 0.8974358974358974, 'agreement': 0.42500000000000004, 'direct_attack': 0.41111111111111115, 'undercutter_attack': 0.33333333333333326, 'partial': 0.14285714285714288}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6578947368421053, 'recall': 0.6578947368421053, 'f1': 0.6578947368421053, 'support': None}, 'macro_avg': {'precision': 0.6306086235489221, 'recall': 0.46010862263457425, 'f1': 0.44194749694749697, 'support': None}, 'weighted_avg': {'precision': 0.6969337522911758, 'recall': 0.6578947368421053, 'f1': 0.6238231797442324, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8035190615835777, 'agreement': 0.4807692307692308, 'direct_attack': 0.2826086956521739, 'undercutter_attack': 0.40229885057471265, 'partial': 0.22727272727272727}, 'recall': {'support': 0.9480968858131488, 'agreement': 0.29411764705882354, 'direct_attack': 0.21666666666666667, 'undercutter_attack': 0.4166666666666667, 'partial': 0.19230769230769232}, 'f1': {'support': 0.8698412698412697, 'agreement': 0.36496350364963503, 'direct_attack': 0.24528301886792453, 'undercutter_attack': 0.4093567251461988, 'partial': 0.20833333333333331}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6263157894736842, 'recall': 0.6263157894736842, 'f1': 0.6263157894736842, 'support': None}, 'macro_avg': {'precision': 0.43929371317048443, 'recall': 0.4135711117025996, 'f1': 0.41955557016767225, 'support': None}, 'weighted_avg': {'precision': 0.5888600007344327, 'recall': 0.6263157894736842, 'f1': 0.60060053363515, 'support': None}}
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8104956268221575, 'agreement': 0.625, 'direct_attack': 0.4117647058823529, 'undercutter_attack': 0.45614035087719296, 'partial': 0.36666666666666664}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.23529411764705882, 'direct_attack': 0.35, 'undercutter_attack': 0.6190476190476191, 'partial': 0.21153846153846154}, 'f1': {'support': 0.8797468354430381, 'agreement': 0.3418803418803419, 'direct_attack': 0.37837837837837834, 'undercutter_attack': 0.5252525252525252, 'partial': 0.26829268292682923}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6701754385964912, 'recall': 0.6701754385964912, 'f1': 0.6701754385964912, 'support': None}, 'macro_avg': {'precision': 0.534013470049674, 'recall': 0.475563582899223, 'f1': 0.4787101527762225, 'support': None}, 'weighted_avg': {'precision': 0.6481518853419219, 'recall': 0.6701754385964912, 'f1': 0.6387399979631175, 'support': None}}
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8211143695014663, 'agreement': 0.625, 'direct_attack': 0.3888888888888889, 'undercutter_attack': 0.43859649122807015, 'partial': 0.2553191489361702}, 'recall': {'support': 0.9688581314878892, 'agreement': 0.23529411764705882, 'direct_attack': 0.23333333333333334, 'undercutter_attack': 0.5952380952380952, 'partial': 0.23076923076923078}, 'f1': {'support': 0.8888888888888891, 'agreement': 0.3418803418803419, 'direct_attack': 0.2916666666666667, 'undercutter_attack': 0.505050505050505, 'partial': 0.24242424242424243}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6596491228070176, 'recall': 0.6596491228070176, 'f1': 0.6596491228070176, 'support': None}, 'macro_avg': {'precision': 0.5057837797109191, 'recall': 0.4526985816951215, 'f1': 0.45398212898212903, 'support': None}, 'weighted_avg': {'precision': 0.6383843633808699, 'recall': 0.6596491228070176, 'f1': 0.6289105631210895, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8372781065088757, 'agreement': 0.5, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.4673913043478261, 'partial': 0.0}, 'recall': {'support': 0.9792387543252595, 'agreement': 0.35294117647058826, 'direct_attack': 0.36666666666666664, 'undercutter_attack': 0.5119047619047619, 'partial': 0.0}, 'f1': {'support': 0.9027113237639554, 'agreement': 0.41379310344827586, 'direct_attack': 0.32116788321167883, 'undercutter_attack': 0.4886363636363636, 'partial': 0.0}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6631578947368421, 'recall': 0.6631578947368421, 'f1': 0.6631578947368421, 'support': None}, 'macro_avg': {'precision': 0.41807673931419753, 'recall': 0.4421502718734553, 'f1': 0.42526173481205476, 'support': None}, 'weighted_avg': {'precision': 0.5980299991037538, 'recall': 0.6631578947368421, 'f1': 0.6252131822790207, 'support': None}}
Loss: tensor(6.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8263473053892215, 'agreement': 0.6052631578947368, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.44776119402985076, 'partial': 0.3055555555555556}, 'recall': {'support': 0.9550173010380623, 'agreement': 0.27058823529411763, 'direct_attack': 0.5, 'undercutter_attack': 0.35714285714285715, 'partial': 0.21153846153846154}, 'f1': {'support': 0.886035313001605, 'agreement': 0.37398373983739835, 'direct_attack': 0.3870967741935484, 'undercutter_attack': 0.3973509933774834, 'partial': 0.25}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6491228070175439, 'recall': 0.6491228070175439, 'f1': 0.6491228070175439, 'support': None}, 'macro_avg': {'precision': 0.5001433373107151, 'recall': 0.45885737100269974, 'f1': 0.458893364082007, 'support': None}, 'weighted_avg': {'precision': 0.6363332233105029, 'recall': 0.6491228070175439, 'f1': 0.6271159881385339, 'support': None}}
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5434, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8279883381924198, 'agreement': 0.5434782608695652, 'direct_attack': 0.4074074074074074, 'undercutter_attack': 0.4953271028037383, 'partial': 0.3}, 'recall': {'support': 0.9826989619377162, 'agreement': 0.29411764705882354, 'direct_attack': 0.36666666666666664, 'undercutter_attack': 0.6309523809523809, 'partial': 0.11538461538461539}, 'f1': {'support': 0.8987341772151899, 'agreement': 0.38167938931297707, 'direct_attack': 0.38596491228070173, 'undercutter_attack': 0.5549738219895288, 'partial': 0.16666666666666669}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6842105263157895, 'recall': 0.6842105263157895, 'f1': 0.6842105263157895, 'support': None}, 'macro_avg': {'precision': 0.5148402218546261, 'recall': 0.47796405440004053, 'f1': 0.47760379349301274, 'support': None}, 'weighted_avg': {'precision': 0.6440986017394402, 'recall': 0.6842105263157895, 'f1': 0.6502092767674073, 'support': None}}
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8528528528528528, 'agreement': 0.8181818181818182, 'direct_attack': 0.3161290322580645, 'undercutter_attack': 0.5555555555555556, 'partial': 0.20689655172413793}, 'recall': {'support': 0.9826989619377162, 'agreement': 0.4235294117647059, 'direct_attack': 0.8166666666666667, 'undercutter_attack': 0.05952380952380952, 'partial': 0.11538461538461539}, 'f1': {'support': 0.9131832797427654, 'agreement': 0.558139534883721, 'direct_attack': 0.455813953488372, 'undercutter_attack': 0.1075268817204301, 'partial': 0.14814814814814817}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6666666666666666, 'recall': 0.6666666666666666, 'f1': 0.6666666666666666, 'support': None}, 'macro_avg': {'precision': 0.5499231621144858, 'recall': 0.4795606930555027, 'f1': 0.43656235959668743, 'support': None}, 'weighted_avg': {'precision': 0.6884437865118153, 'recall': 0.6666666666666666, 'f1': 0.6235730303303466, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(11.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8503184713375797, 'agreement': 0.4, 'direct_attack': 0.35555555555555557, 'undercutter_attack': 0.46551724137931033, 'partial': 0.32}, 'recall': {'support': 0.9238754325259516, 'agreement': 0.32941176470588235, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.6428571428571429, 'partial': 0.15384615384615385}, 'f1': {'support': 0.8855721393034826, 'agreement': 0.36129032258064514, 'direct_attack': 0.3047619047619048, 'undercutter_attack': 0.5399999999999999, 'partial': 0.2077922077922078}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6543859649122807, 'recall': 0.6543859649122807, 'f1': 0.6543859649122807, 'support': None}, 'macro_avg': {'precision': 0.4782782536544891, 'recall': 0.4633314321203595, 'f1': 0.4598833148876481, 'support': None}, 'weighted_avg': {'precision': 0.6259979295188701, 'recall': 0.6543859649122807, 'f1': 0.6334928680157376, 'support': None}}
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8114285714285714, 'agreement': 0.46938775510204084, 'direct_attack': 0.3142857142857143, 'undercutter_attack': 0.4375, 'partial': 0.2222222222222222}, 'recall': {'support': 0.9826989619377162, 'agreement': 0.27058823529411763, 'direct_attack': 0.55, 'undercutter_attack': 0.25, 'partial': 0.07692307692307693}, 'f1': {'support': 0.8888888888888888, 'agreement': 0.3432835820895522, 'direct_attack': 0.4, 'undercutter_attack': 0.3181818181818182, 'partial': 0.1142857142857143}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6403508771929824, 'recall': 0.6403508771929824, 'f1': 0.6403508771929824, 'support': None}, 'macro_avg': {'precision': 0.4509648526077098, 'recall': 0.4260420548309821, 'f1': 0.4129280006891946, 'support': None}, 'weighted_avg': {'precision': 0.5992342363846123, 'recall': 0.6403508771929824, 'f1': 0.6012949530467204, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8221574344023324, 'agreement': 0.6, 'direct_attack': 0.32653061224489793, 'undercutter_attack': 0.41346153846153844, 'partial': 0.4166666666666667}, 'recall': {'support': 0.9757785467128027, 'agreement': 0.35294117647058826, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.5119047619047619, 'partial': 0.19230769230769232}, 'f1': {'support': 0.8924050632911391, 'agreement': 0.4444444444444445, 'direct_attack': 0.29357798165137616, 'undercutter_attack': 0.4574468085106383, 'partial': 0.2631578947368421}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6684210526315789, 'recall': 0.6684210526315789, 'f1': 0.6684210526315789, 'support': None}, 'macro_avg': {'precision': 0.515763250355087, 'recall': 0.4599197688125024, 'f1': 0.470206438526888, 'support': None}, 'weighted_avg': {'precision': 0.6396364406568488, 'recall': 0.6684210526315789, 'f1': 0.6410653726477349, 'support': None}}
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2489e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.889273356401384, 'agreement': 0.4659090909090909, 'direct_attack': 0.29518072289156627, 'undercutter_attack': 0.6, 'partial': 0.25}, 'recall': {'support': 0.889273356401384, 'agreement': 0.4823529411764706, 'direct_attack': 0.8166666666666667, 'undercutter_attack': 0.10714285714285714, 'partial': 0.057692307692307696}, 'f1': {'support': 0.889273356401384, 'agreement': 0.47398843930635837, 'direct_attack': 0.43362831858407086, 'undercutter_attack': 0.18181818181818182, 'partial': 0.09375}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6298245614035087, 'recall': 0.6298245614035087, 'f1': 0.6298245614035087, 'support': None}, 'macro_avg': {'precision': 0.5000726340404082, 'recall': 0.4706256258159372, 'f1': 0.41449165922199904, 'support': None}, 'weighted_avg': {'precision': 0.6626545896504679, 'recall': 0.6298245614035087, 'f1': 0.6025516556645824, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.907258064516129, 'agreement': 0.3826086956521739, 'direct_attack': 0.3048780487804878, 'undercutter_attack': 0.5, 'partial': 0.19402985074626866}, 'recall': {'support': 0.7785467128027682, 'agreement': 0.5176470588235295, 'direct_attack': 0.4166666666666667, 'undercutter_attack': 0.34523809523809523, 'partial': 0.25}, 'f1': {'support': 0.8379888268156424, 'agreement': 0.44, 'direct_attack': 0.35211267605633806, 'undercutter_attack': 0.4084507042253521, 'partial': 0.2184873949579832}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.5894736842105263, 'recall': 0.5894736842105263, 'f1': 0.5894736842105263, 'support': None}, 'macro_avg': {'precision': 0.4577549319390119, 'recall': 0.4616197067062119, 'f1': 0.45140792041106315, 'support': None}, 'weighted_avg': {'precision': 0.6405290437565463, 'recall': 0.5894736842105263, 'f1': 0.6076784828172731, 'support': None}}
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8806451612903226, 'agreement': 0.6290322580645161, 'direct_attack': 0.34285714285714286, 'undercutter_attack': 0.5066666666666667, 'partial': 0.2641509433962264}, 'recall': {'support': 0.9446366782006921, 'agreement': 0.4588235294117647, 'direct_attack': 0.4, 'undercutter_attack': 0.4523809523809524, 'partial': 0.2692307692307692}, 'f1': {'support': 0.9115191986644408, 'agreement': 0.5306122448979591, 'direct_attack': 0.36923076923076925, 'undercutter_attack': 0.47798742138364775, 'partial': 0.26666666666666666}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.524670434454975, 'recall': 0.5050143858448357, 'f1': 0.5112032601686967, 'support': None}, 'weighted_avg': {'precision': 0.6751604757481043, 'recall': 0.6807017543859649, 'f1': 0.6749167463984019, 'support': None}}
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3852e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8575851393188855, 'agreement': 0.6, 'direct_attack': 0.34065934065934067, 'undercutter_attack': 0.5, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9584775086505191, 'agreement': 0.4235294117647059, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.35714285714285715, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9052287581699346, 'agreement': 0.496551724137931, 'direct_attack': 0.4105960264900662, 'undercutter_attack': 0.41666666666666663, 'partial': 0.2272727272727273}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6736842105263158, 'recall': 0.6736842105263158, 'f1': 0.6736842105263158, 'support': None}, 'macro_avg': {'precision': 0.5152044515512008, 'recall': 0.48962482730648826, 'f1': 0.49126318054746515, 'support': None}, 'weighted_avg': {'precision': 0.6591686142932681, 'recall': 0.6736842105263158, 'f1': 0.6583718439831947, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6294e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.64, 'direct_attack': 0.36585365853658536, 'undercutter_attack': 0.5277777777777778, 'partial': 0.2631578947368421}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5, 'undercutter_attack': 0.4523809523809524, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47407407407407404, 'direct_attack': 0.4225352112676056, 'undercutter_attack': 0.4871794871794871, 'partial': 0.2222222222222222}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.5288700613321923, 'recall': 0.496619389837383, 'f1': 0.5014291033246908, 'support': None}, 'weighted_avg': {'precision': 0.6654629567071293, 'recall': 0.6807017543859649, 'f1': 0.6641314355689891, 'support': None}}
Loss: tensor(8.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.64, 'direct_attack': 0.36470588235294116, 'undercutter_attack': 0.5507246376811594, 'partial': 0.2631578947368421}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.4523809523809524, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47407407407407404, 'direct_attack': 0.42758620689655175, 'undercutter_attack': 0.49673202614379086, 'partial': 0.2222222222222222}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6824561403508772, 'recall': 0.6824561403508772, 'f1': 0.6824561403508772, 'support': None}, 'macro_avg': {'precision': 0.5332298780761398, 'recall': 0.4999527231707163, 'f1': 0.5043498102433407, 'support': None}, 'weighted_avg': {'precision': 0.6687237806735599, 'recall': 0.6824561403508772, 'f1': 0.6660708619036178, 'support': None}}
Loss: tensor(6.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.6274509803921569, 'direct_attack': 0.36046511627906974, 'undercutter_attack': 0.5441176470588235, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47058823529411764, 'direct_attack': 0.4246575342465753, 'undercutter_attack': 0.4868421052631579, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.5299729979220154, 'recall': 0.4975717707897639, 'f1': 0.5015882995615023, 'support': None}, 'weighted_avg': {'precision': 0.666081225733846, 'recall': 0.6807017543859649, 'f1': 0.6640130859274834, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8135e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.6274509803921569, 'direct_attack': 0.36046511627906974, 'undercutter_attack': 0.5441176470588235, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47058823529411764, 'direct_attack': 0.4246575342465753, 'undercutter_attack': 0.4868421052631579, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.5299729979220154, 'recall': 0.4975717707897639, 'f1': 0.5015882995615023, 'support': None}, 'weighted_avg': {'precision': 0.666081225733846, 'recall': 0.6807017543859649, 'f1': 0.6640130859274834, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7808e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.6274509803921569, 'direct_attack': 0.36046511627906974, 'undercutter_attack': 0.5441176470588235, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47058823529411764, 'direct_attack': 0.4246575342465753, 'undercutter_attack': 0.4868421052631579, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.5299729979220154, 'recall': 0.4975717707897639, 'f1': 0.5015882995615023, 'support': None}, 'weighted_avg': {'precision': 0.666081225733846, 'recall': 0.6807017543859649, 'f1': 0.6640130859274834, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9338e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6605e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4772e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.6274509803921569, 'direct_attack': 0.3563218390804598, 'undercutter_attack': 0.5522388059701493, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47058823529411764, 'direct_attack': 0.42176870748299317, 'undercutter_attack': 0.4900662251655629, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.5307685742645585, 'recall': 0.4975717707897639, 'f1': 0.5016553581892669, 'support': None}, 'weighted_avg': {'precision': 0.6668418936577666, 'recall': 0.6807017543859649, 'f1': 0.6641841323590397, 'support': None}}
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4780e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3569e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8475609756097561, 'agreement': 0.6274509803921569, 'direct_attack': 0.3563218390804598, 'undercutter_attack': 0.5522388059701493, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.3764705882352941, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9011345218800648, 'agreement': 0.47058823529411764, 'direct_attack': 0.42176870748299317, 'undercutter_attack': 0.4900662251655629, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6807017543859649, 'recall': 0.6807017543859649, 'f1': 0.6807017543859649, 'support': None}, 'macro_avg': {'precision': 0.5307685742645585, 'recall': 0.4975717707897639, 'f1': 0.5016553581892669, 'support': None}, 'weighted_avg': {'precision': 0.6668418936577666, 'recall': 0.6807017543859649, 'f1': 0.6641841323590397, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6559e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9665e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8501529051987767, 'agreement': 0.6346153846153846, 'direct_attack': 0.3563218390804598, 'undercutter_attack': 0.5522388059701493, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.38823529411764707, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9025974025974025, 'agreement': 0.48175182481751827, 'direct_attack': 0.42176870748299317, 'undercutter_attack': 0.4900662251655629, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6824561403508772, 'recall': 0.6824561403508772, 'f1': 0.6824561403508772, 'support': None}, 'macro_avg': {'precision': 0.5327198410270081, 'recall': 0.49992471196623445, 'f1': 0.5041806522374145, 'support': None}, 'weighted_avg': {'precision': 0.6692244235002253, 'recall': 0.6824561403508772, 'f1': 0.666590584353425, 'support': None}}
Loss: tensor(4.5037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7525e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5091e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7875e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0320e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8481e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7087e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4780e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7270e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8708e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7628e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0852e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8501529051987767, 'agreement': 0.6346153846153846, 'direct_attack': 0.36046511627906974, 'undercutter_attack': 0.5441176470588235, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.38823529411764707, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9025974025974025, 'agreement': 0.48175182481751827, 'direct_attack': 0.4246575342465753, 'undercutter_attack': 0.4868421052631579, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6824561403508772, 'recall': 0.6824561403508772, 'f1': 0.6824561403508772, 'support': None}, 'macro_avg': {'precision': 0.531924264684465, 'recall': 0.49992471196623445, 'f1': 0.5041135936096499, 'support': None}, 'weighted_avg': {'precision': 0.6684637555763046, 'recall': 0.6824561403508772, 'f1': 0.6664195379218687, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2055e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9665e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5344e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8973e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9473e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3896e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3896e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6893e-07, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7290e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8501529051987767, 'agreement': 0.6346153846153846, 'direct_attack': 0.36046511627906974, 'undercutter_attack': 0.5441176470588235, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9619377162629758, 'agreement': 0.38823529411764707, 'direct_attack': 0.5166666666666667, 'undercutter_attack': 0.44047619047619047, 'partial': 0.19230769230769232}, 'f1': {'support': 0.9025974025974025, 'agreement': 0.48175182481751827, 'direct_attack': 0.4246575342465753, 'undercutter_attack': 0.4868421052631579, 'partial': 0.22471910112359553}, 'support': {'support': 289, 'agreement': 85, 'direct_attack': 60, 'undercutter_attack': 84, 'partial': 52}, 'micro_avg': {'precision': 0.6824561403508772, 'recall': 0.6824561403508772, 'f1': 0.6824561403508772, 'support': None}, 'macro_avg': {'precision': 0.531924264684465, 'recall': 0.49992471196623445, 'f1': 0.5041135936096499, 'support': None}, 'weighted_avg': {'precision': 0.6684637555763046, 'recall': 0.6824561403508772, 'f1': 0.6664195379218687, 'support': None}}


		-------------RUN 3-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.15254237288135594, 'undercutter_attack': 0.375, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 1.0, 'undercutter_attack': 0.03571428571428571, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.2647058823529412, 'undercutter_attack': 0.06521739130434782, 'partial': 0.0}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.15625, 'recall': 0.15625, 'f1': 0.15625, 'support': None}, 'macro_avg': {'precision': 0.10550847457627119, 'recall': 0.20714285714285716, 'f1': 0.0659846547314578, 'support': None}, 'weighted_avg': {'precision': 0.08850635593220339, 'recall': 0.15625, 'f1': 0.05111892583120204, 'support': None}}
Loss: tensor(3.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9771, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8817204301075269, 'agreement': 0.5, 'direct_attack': 0.0, 'undercutter_attack': 0.3, 'partial': 0.07142857142857142}, 'recall': {'support': 0.703862660944206, 'agreement': 0.2702702702702703, 'direct_attack': 0.0, 'undercutter_attack': 0.9285714285714286, 'partial': 0.018518518518518517}, 'f1': {'support': 0.782816229116945, 'agreement': 0.3508771929824562, 'direct_attack': 0.0, 'undercutter_attack': 0.45348837209302323, 'partial': 0.029411764705882353}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.5270833333333333, 'recall': 0.5270833333333333, 'f1': 0.5270833333333333, 'support': None}, 'macro_avg': {'precision': 0.3506298003072197, 'recall': 0.3842445756608847, 'f1': 0.3233187117796613, 'support': None}, 'weighted_avg': {'precision': 0.5270791730670763, 'recall': 0.5270833333333333, 'f1': 0.48970811682193893, 'support': None}}
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0965, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7597173144876325, 'agreement': 0.45714285714285713, 'direct_attack': 0.1724137931034483, 'undercutter_attack': 0.32673267326732675, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9227467811158798, 'agreement': 0.43243243243243246, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.39285714285714285, 'partial': 0.018518518518518517}, 'f1': {'support': 0.8333333333333334, 'agreement': 0.4444444444444445, 'direct_attack': 0.15384615384615385, 'undercutter_attack': 0.3567567567567567, 'partial': 0.03508771929824561}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.5729166666666666, 'recall': 0.5729166666666666, 'f1': 0.5729166666666666, 'support': None}, 'macro_avg': {'precision': 0.40986799426691956, 'recall': 0.3810887527625725, 'f1': 0.36469368153578685, 'support': None}, 'weighted_avg': {'precision': 0.524557828432933, 'recall': 0.5729166666666666, 'f1': 0.5282298720785563, 'support': None}}
Loss: tensor(1.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.7122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9385964912280702, 'agreement': 0.3404255319148936, 'direct_attack': 0.2708333333333333, 'undercutter_attack': 0.36470588235294116, 'partial': 0.4166666666666667}, 'recall': {'support': 0.9145299145299145, 'agreement': 0.43243243243243246, 'direct_attack': 0.36619718309859156, 'undercutter_attack': 0.36904761904761907, 'partial': 0.18518518518518517}, 'f1': {'support': 0.9264069264069265, 'agreement': 0.38095238095238093, 'direct_attack': 0.31137724550898205, 'undercutter_attack': 0.3668639053254437, 'partial': 0.2564102564102564}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.61875, 'recall': 0.61875, 'f1': 0.61875, 'support': None}, 'macro_avg': {'precision': 0.466245581099181, 'recall': 0.4534784668587485, 'f1': 0.4484021429207979, 'support': None}, 'weighted_avg': {'precision': 0.6345662175261109, 'recall': 0.61875, 'f1': 0.6200936774980995, 'support': None}}
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8566037735849057, 'agreement': 0.38235294117647056, 'direct_attack': 0.25, 'undercutter_attack': 0.4195804195804196, 'partial': 0.5}, 'recall': {'support': 0.9700854700854701, 'agreement': 0.3611111111111111, 'direct_attack': 0.09722222222222222, 'undercutter_attack': 0.7142857142857143, 'partial': 0.09259259259259259}, 'f1': {'support': 0.909819639278557, 'agreement': 0.3714285714285714, 'direct_attack': 0.14, 'undercutter_attack': 0.5286343612334802, 'partial': 0.15625}, 'support': {'support': 234, 'agreement': 36, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.65, 'recall': 0.65, 'f1': 0.65, 'support': None}, 'macro_avg': {'precision': 0.48170742686835916, 'recall': 0.44705942205942206, 'f1': 0.42122651438812175, 'support': None}, 'weighted_avg': {'precision': 0.6134473836374502, 'recall': 0.65, 'f1': 0.6024833552212985, 'support': None}}
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8911290322580645, 'agreement': 0.40625, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.44, 'partial': 1.0}, 'recall': {'support': 0.9484978540772532, 'agreement': 0.7027027027027027, 'direct_attack': 0.2222222222222222, 'undercutter_attack': 0.5238095238095238, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9189189189189189, 'agreement': 0.5148514851485149, 'direct_attack': 0.2318840579710145, 'undercutter_attack': 0.4782608695652174, 'partial': 0.07142857142857142}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.64375, 'recall': 0.64375, 'f1': 0.64375, 'support': None}, 'macro_avg': {'precision': 0.5959606549364613, 'recall': 0.4868538679697478, 'f1': 0.44306878060644744, 'support': None}, 'weighted_avg': {'precision': 0.6897476249389051, 'recall': 0.64375, 'f1': 0.6122590023607027, 'support': None}}
Loss: tensor(0.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8488372093023255, 'agreement': 0.26595744680851063, 'direct_attack': 0.20833333333333334, 'undercutter_attack': 0.47126436781609193, 'partial': 0.23529411764705882}, 'recall': {'support': 0.9399141630901288, 'agreement': 0.6756756756756757, 'direct_attack': 0.06944444444444445, 'undercutter_attack': 0.4880952380952381, 'partial': 0.07407407407407407}, 'f1': {'support': 0.8920570264765785, 'agreement': 0.38167938931297707, 'direct_attack': 0.10416666666666667, 'undercutter_attack': 0.47953216374269, 'partial': 0.11267605633802816}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6125, 'recall': 0.6125, 'f1': 0.6125, 'support': None}, 'macro_avg': {'precision': 0.405937294981464, 'recall': 0.4494407190759121, 'f1': 0.3940222605073881, 'support': None}, 'weighted_avg': {'precision': 0.5727324678101035, 'recall': 0.6125, 'f1': 0.5746596528547133, 'support': None}}
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.88, 'agreement': 0.28735632183908044, 'direct_attack': 0.25, 'undercutter_attack': 0.47619047619047616, 'partial': 0.21428571428571427}, 'recall': {'support': 0.944206008583691, 'agreement': 0.6756756756756757, 'direct_attack': 0.08333333333333333, 'undercutter_attack': 0.5952380952380952, 'partial': 0.05555555555555555}, 'f1': {'support': 0.9109730848861284, 'agreement': 0.4032258064516129, 'direct_attack': 0.125, 'undercutter_attack': 0.5291005291005291, 'partial': 0.08823529411764706}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6333333333333333, 'recall': 0.6333333333333333, 'f1': 0.6333333333333333, 'support': None}, 'macro_avg': {'precision': 0.42156650246305427, 'recall': 0.47080173367727013, 'f1': 0.4113069429111834, 'support': None}, 'weighted_avg': {'precision': 0.5942575259989054, 'recall': 0.6333333333333333, 'f1': 0.5945525707166144, 'support': None}}
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8599221789883269, 'agreement': 0.29577464788732394, 'direct_attack': 0.21739130434782608, 'undercutter_attack': 0.4536082474226804, 'partial': 0.2222222222222222}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5675675675675675, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5301204819277109, 'partial': 0.037037037037037035}, 'f1': {'support': 0.90020366598778, 'agreement': 0.38888888888888884, 'direct_attack': 0.1694915254237288, 'undercutter_attack': 0.4888888888888889, 'partial': 0.06349206349206349}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 83, 'partial': 54}, 'micro_avg': {'precision': 0.6208333333333333, 'recall': 0.6208333333333333, 'f1': 0.6208333333333333, 'support': None}, 'macro_avg': {'precision': 0.4097837201736759, 'recall': 0.4436116839731298, 'f1': 0.40219300653626994, 'support': None}, 'weighted_avg': {'precision': 0.5780564798004696, 'recall': 0.6208333333333333, 'f1': 0.5859297620143482, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.31343283582089554, 'direct_attack': 0.25, 'undercutter_attack': 0.43636363636363634, 'partial': 0.2}, 'recall': {'support': 0.9401709401709402, 'agreement': 0.5675675675675675, 'direct_attack': 0.15492957746478872, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9109730848861283, 'agreement': 0.40384615384615385, 'direct_attack': 0.19130434782608693, 'undercutter_attack': 0.4948453608247423, 'partial': 0.0625}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6291666666666667, 'recall': 0.6291666666666667, 'f1': 0.6291666666666667, 'support': None}, 'macro_avg': {'precision': 0.4166661217461433, 'recall': 0.45422673873378105, 'f1': 0.4126937894766223, 'support': None}, 'weighted_avg': {'precision': 0.5907261423577622, 'recall': 0.6291666666666667, 'f1': 0.5971554761679004, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.88, 'agreement': 0.31343283582089554, 'direct_attack': 0.25, 'undercutter_attack': 0.42727272727272725, 'partial': 0.2222222222222222}, 'recall': {'support': 0.9401709401709402, 'agreement': 0.5833333333333334, 'direct_attack': 0.1527777777777778, 'undercutter_attack': 0.5595238095238095, 'partial': 0.037037037037037035}, 'f1': {'support': 0.909090909090909, 'agreement': 0.4077669902912622, 'direct_attack': 0.18965517241379312, 'undercutter_attack': 0.4845360824742268, 'partial': 0.06349206349206349}, 'support': {'support': 234, 'agreement': 36, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.418585557063169, 'recall': 0.45456857956857954, 'f1': 0.410908243552451, 'support': None}, 'weighted_avg': {'precision': 0.5897801899592944, 'recall': 0.6270833333333333, 'f1': 0.5941492898915787, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8804780876494024, 'agreement': 0.31343283582089554, 'direct_attack': 0.2619047619047619, 'undercutter_attack': 0.43636363636363634, 'partial': 0.2}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5675675675675675, 'direct_attack': 0.1527777777777778, 'undercutter_attack': 0.5783132530120482, 'partial': 0.037037037037037035}, 'f1': {'support': 0.911340206185567, 'agreement': 0.40384615384615385, 'direct_attack': 0.1929824561403509, 'undercutter_attack': 0.4974093264248704, 'partial': 0.0625}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 83, 'partial': 54}, 'micro_avg': {'precision': 0.63125, 'recall': 0.63125, 'f1': 0.63125, 'support': None}, 'macro_avg': {'precision': 0.41843586434773916, 'recall': 0.4560280159677751, 'f1': 0.41361562851938843, 'support': None}, 'weighted_avg': {'precision': 0.5906337752305375, 'recall': 0.63125, 'f1': 0.5973971393231247, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8804780876494024, 'agreement': 0.3235294117647059, 'direct_attack': 0.2558139534883721, 'undercutter_attack': 0.43636363636363634, 'partial': 0.25}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5945945945945946, 'direct_attack': 0.15492957746478872, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.911340206185567, 'agreement': 0.41904761904761906, 'direct_attack': 0.19298245614035084, 'undercutter_attack': 0.4948453608247423, 'partial': 0.06451612903225806}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6333333333333333, 'recall': 0.6333333333333333, 'f1': 0.6333333333333333, 'support': None}, 'macro_avg': {'precision': 0.4292370178532233, 'recall': 0.46048684499388726, 'f1': 0.4165463542461074, 'support': None}, 'weighted_avg': {'precision': 0.5964995768697378, 'recall': 0.6333333333333333, 'f1': 0.598981262114937, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8875502008032129, 'agreement': 0.3235294117647059, 'direct_attack': 0.2558139534883721, 'undercutter_attack': 0.43243243243243246, 'partial': 0.2222222222222222}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5945945945945946, 'direct_attack': 0.15492957746478872, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9151138716356108, 'agreement': 0.41904761904761906, 'direct_attack': 0.19298245614035084, 'undercutter_attack': 0.49230769230769234, 'partial': 0.06349206349206349}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6333333333333333, 'recall': 0.6333333333333333, 'f1': 0.6333333333333333, 'support': None}, 'macro_avg': {'precision': 0.42430964414218914, 'recall': 0.46048684499388726, 'f1': 0.4165887405246673, 'support': None}, 'weighted_avg': {'precision': 0.5961342713442598, 'recall': 0.6333333333333333, 'f1': 0.6002616246580778, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8906882591093117, 'agreement': 0.31343283582089554, 'direct_attack': 0.24390243902439024, 'undercutter_attack': 0.41739130434782606, 'partial': 0.2}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5675675675675675, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9166666666666666, 'agreement': 0.40384615384615385, 'direct_attack': 0.1769911504424779, 'undercutter_attack': 0.48241206030150746, 'partial': 0.0625}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.4130829676604847, 'recall': 0.4518256147011511, 'f1': 0.40848320625136114, 'support': None}, 'weighted_avg': {'precision': 0.5886442176517005, 'recall': 0.6270833333333333, 'f1': 0.594097118589221, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8911290322580645, 'agreement': 0.3181818181818182, 'direct_attack': 0.24390243902439024, 'undercutter_attack': 0.41739130434782606, 'partial': 0.2}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5675675675675675, 'direct_attack': 0.14084507042253522, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9170124481327802, 'agreement': 0.4077669902912622, 'direct_attack': 0.17857142857142855, 'undercutter_attack': 0.48241206030150746, 'partial': 0.0625}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6291666666666667, 'recall': 0.6291666666666667, 'f1': 0.6291666666666667, 'support': None}, 'macro_avg': {'precision': 0.4141209187624198, 'recall': 0.4522645381800311, 'f1': 0.40965258545939565, 'support': None}, 'weighted_avg': {'precision': 0.5905726324105489, 'recall': 0.6291666666666667, 'f1': 0.5963426583286361, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8906882591093117, 'agreement': 0.31343283582089554, 'direct_attack': 0.24390243902439024, 'undercutter_attack': 0.41379310344827586, 'partial': 0.2222222222222222}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5675675675675675, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9166666666666666, 'agreement': 0.40384615384615385, 'direct_attack': 0.1769911504424779, 'undercutter_attack': 0.48000000000000004, 'partial': 0.06349206349206349}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.4168077719250191, 'recall': 0.4518256147011511, 'f1': 0.4081992068894724, 'support': None}, 'weighted_avg': {'precision': 0.5905145324942792, 'recall': 0.6270833333333333, 'f1': 0.5937866151793142, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8870967741935484, 'agreement': 0.3181818181818182, 'direct_attack': 0.25, 'undercutter_attack': 0.4067796610169492, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5675675675675675, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5714285714285714, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9147609147609147, 'agreement': 0.4077669902912622, 'direct_attack': 0.17857142857142858, 'undercutter_attack': 0.4752475247524753, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42241165067846315, 'recall': 0.4518256147011511, 'f1': 0.4081725974816678, 'support': None}, 'weighted_avg': {'precision': 0.5919495149692663, 'recall': 0.6270833333333333, 'f1': 0.5926843285086719, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8875502008032129, 'agreement': 0.3230769230769231, 'direct_attack': 0.2564102564102564, 'undercutter_attack': 0.4117647058823529, 'partial': 0.25}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5675675675675675, 'direct_attack': 0.14084507042253522, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9151138716356108, 'agreement': 0.411764705882353, 'direct_attack': 0.1818181818181818, 'undercutter_attack': 0.4827586206896552, 'partial': 0.06451612903225806}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.63125, 'recall': 0.63125, 'f1': 0.63125, 'support': None}, 'macro_avg': {'precision': 0.42576041723454905, 'recall': 0.4546454905609836, 'f1': 0.41119430181161176, 'support': None}, 'weighted_avg': {'precision': 0.5956957430021745, 'recall': 0.63125, 'f1': 0.5964929710315497, 'support': None}}
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8870967741935484, 'agreement': 0.3181818181818182, 'direct_attack': 0.2564102564102564, 'undercutter_attack': 0.4117647058823529, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5675675675675675, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9147609147609147, 'agreement': 0.4077669902912622, 'direct_attack': 0.1801801801801802, 'undercutter_attack': 0.4827586206896552, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6291666666666667, 'recall': 0.6291666666666667, 'f1': 0.6291666666666667, 'support': None}, 'macro_avg': {'precision': 0.42469071093359523, 'recall': 0.4542065670821036, 'f1': 0.409996566990854, 'support': None}, 'weighted_avg': {'precision': 0.5937834362822503, 'recall': 0.6291666666666667, 'f1': 0.5942400830389912, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8870967741935484, 'agreement': 0.3181818181818182, 'direct_attack': 0.2564102564102564, 'undercutter_attack': 0.4117647058823529, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5675675675675675, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9147609147609147, 'agreement': 0.4077669902912622, 'direct_attack': 0.1801801801801802, 'undercutter_attack': 0.4827586206896552, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6291666666666667, 'recall': 0.6291666666666667, 'f1': 0.6291666666666667, 'support': None}, 'macro_avg': {'precision': 0.42469071093359523, 'recall': 0.4542065670821036, 'f1': 0.409996566990854, 'support': None}, 'weighted_avg': {'precision': 0.5937834362822503, 'recall': 0.6291666666666667, 'f1': 0.5942400830389912, 'support': None}}
Loss: tensor(9.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2047e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.884, 'agreement': 0.3181818181818182, 'direct_attack': 0.2631578947368421, 'undercutter_attack': 0.4152542372881356, 'partial': 0.25}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5675675675675675, 'direct_attack': 0.14084507042253522, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9132231404958677, 'agreement': 0.4077669902912622, 'direct_attack': 0.1834862385321101, 'undercutter_attack': 0.4851485148514852, 'partial': 0.06451612903225806}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.63125, 'recall': 0.63125, 'f1': 0.63125, 'support': None}, 'macro_avg': {'precision': 0.4261187900413592, 'recall': 0.4546454905609836, 'f1': 0.4108282026405966, 'support': None}, 'weighted_avg': {'precision': 0.5951964452734301, 'recall': 0.63125, 'f1': 0.5959280472247006, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2047e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.30303030303030304, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9128630705394192, 'agreement': 0.38834951456310685, 'direct_attack': 0.18348623853211007, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42303360863601824, 'recall': 0.44880116167669815, 'f1': 0.4059214219059279, 'support': None}, 'weighted_avg': {'precision': 0.5923646551809203, 'recall': 0.6270833333333333, 'f1': 0.5919038516521752, 'support': None}}
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.30303030303030304, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9128630705394192, 'agreement': 0.38834951456310685, 'direct_attack': 0.18348623853211007, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42303360863601824, 'recall': 0.44880116167669815, 'f1': 0.4059214219059279, 'support': None}, 'weighted_avg': {'precision': 0.5923646551809203, 'recall': 0.6270833333333333, 'f1': 0.5919038516521752, 'support': None}}
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.884, 'agreement': 0.30303030303030304, 'direct_attack': 0.25, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5405405405405406, 'direct_attack': 0.1267605633802817, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9132231404958677, 'agreement': 0.38834951456310685, 'direct_attack': 0.16822429906542055, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 234, 'agreement': 37, 'direct_attack': 71, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.41907272727272726, 'recall': 0.4464231837471274, 'f1': 0.40294104800387964, 'support': None}, 'weighted_avg': {'precision': 0.5908710858585858, 'recall': 0.6270833333333333, 'f1': 0.5913414256098445, 'support': None}}
Loss: tensor(7.4791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.30303030303030304, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9128630705394192, 'agreement': 0.38834951456310685, 'direct_attack': 0.18348623853211007, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42303360863601824, 'recall': 0.44880116167669815, 'f1': 0.4059214219059279, 'support': None}, 'weighted_avg': {'precision': 0.5923646551809203, 'recall': 0.6270833333333333, 'f1': 0.5919038516521752, 'support': None}}
Loss: tensor(3.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4772e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9665e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1246e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9432e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0852e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9113e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0902e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.30303030303030304, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9128630705394192, 'agreement': 0.38834951456310685, 'direct_attack': 0.18348623853211007, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42303360863601824, 'recall': 0.44880116167669815, 'f1': 0.4059214219059279, 'support': None}, 'weighted_avg': {'precision': 0.5923646551809203, 'recall': 0.6270833333333333, 'f1': 0.5919038516521752, 'support': None}}
Loss: tensor(1.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2760e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4914e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4321e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9202e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7816e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5208e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9170e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3896e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8854e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.30303030303030304, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9128630705394192, 'agreement': 0.38834951456310685, 'direct_attack': 0.18348623853211007, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42303360863601824, 'recall': 0.44880116167669815, 'f1': 0.4059214219059279, 'support': None}, 'weighted_avg': {'precision': 0.5923646551809203, 'recall': 0.6270833333333333, 'f1': 0.5919038516521752, 'support': None}}
Loss: tensor(1.4109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6743e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7022e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7942e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7308e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3671e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7302e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9330e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8599e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5782e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1258e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8201e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1264e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9665e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7353e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2751e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8224e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6605e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3896e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5571e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1279e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8835341365461847, 'agreement': 0.30303030303030304, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.4083333333333333, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.5833333333333334, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9128630705394192, 'agreement': 0.38834951456310685, 'direct_attack': 0.18348623853211007, 'undercutter_attack': 0.4803921568627451, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6270833333333333, 'recall': 0.6270833333333333, 'f1': 0.6270833333333333, 'support': None}, 'macro_avg': {'precision': 0.42303360863601824, 'recall': 0.44880116167669815, 'f1': 0.4059214219059279, 'support': None}, 'weighted_avg': {'precision': 0.5923646551809203, 'recall': 0.6270833333333333, 'f1': 0.5919038516521752, 'support': None}}
Loss: tensor(2.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5228e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6707e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8792e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5473e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0958e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7915e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1249e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6465e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7593e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2783e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5832e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9505e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5170e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7808e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0852e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4372e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9493e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2460e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2369e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8256e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5571e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4393e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5229e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4891e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6102e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0098e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7688e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9799e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2786e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9818e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3714e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6975e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9176e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5188e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7703e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7930e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8875502008032129, 'agreement': 0.2923076923076923, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.42276422764227645, 'partial': 0.25}, 'recall': {'support': 0.9444444444444444, 'agreement': 0.5277777777777778, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.6190476190476191, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9151138716356108, 'agreement': 0.3762376237623763, 'direct_attack': 0.18691588785046728, 'undercutter_attack': 0.502415458937198, 'partial': 0.06451612903225806}, 'support': {'support': 234, 'agreement': 36, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6333333333333333, 'recall': 0.6333333333333333, 'f1': 0.6333333333333333, 'support': None}, 'macro_avg': {'precision': 0.4276672812934935, 'recall': 0.4534391534391535, 'f1': 0.40903979424358206, 'support': None}, 'weighted_avg': {'precision': 0.5995696825091844, 'recall': 0.6333333333333333, 'f1': 0.5975539872122473, 'support': None}}
Loss: tensor(6.5808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9499e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5803e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1883e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3720e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0691e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6613e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9994e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3401e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7505e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9726e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4337e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4055e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4931e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8934e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6094e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7587e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1261e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4357e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2859e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5540e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5467e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2192e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5871e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7353e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7070e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8507e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0679e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3982e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0070e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8282e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6206e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9657e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3752e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1903e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8851e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8070e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5756e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0145e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5091e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8212e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6091e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4360e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2047e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1033e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5264e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9429e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4931e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4971e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5806e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7942e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1635e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0704e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0961e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9670e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8870967741935484, 'agreement': 0.30303030303030304, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.42276422764227645, 'partial': 0.25}, 'recall': {'support': 0.944206008583691, 'agreement': 0.5405405405405406, 'direct_attack': 0.1388888888888889, 'undercutter_attack': 0.6190476190476191, 'partial': 0.037037037037037035}, 'f1': {'support': 0.9147609147609147, 'agreement': 0.38834951456310685, 'direct_attack': 0.18691588785046728, 'undercutter_attack': 0.502415458937198, 'partial': 0.06451612903225806}, 'support': {'support': 233, 'agreement': 37, 'direct_attack': 72, 'undercutter_attack': 84, 'partial': 54}, 'micro_avg': {'precision': 0.6333333333333333, 'recall': 0.6333333333333333, 'f1': 0.6333333333333333, 'support': None}, 'macro_avg': {'precision': 0.42972111811608277, 'recall': 0.4559440188195554, 'f1': 0.411391581028789, 'support': None}, 'weighted_avg': {'precision': 0.598936027692912, 'recall': 0.6333333333333333, 'f1': 0.597193622128809, 'support': None}}


		-------------RUN 4-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.10516605166051661, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 1.0, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.1903171953255426, 'partial': 0.0}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 44}, 'micro_avg': {'precision': 0.10516605166051661, 'recall': 0.10516605166051661, 'f1': 0.10516605166051661, 'support': None}, 'macro_avg': {'precision': 0.02103321033210332, 'recall': 0.2, 'f1': 0.03806343906510852, 'support': None}, 'weighted_avg': {'precision': 0.011059898421862448, 'recall': 0.10516605166051661, 'f1': 0.02001490799549064, 'support': None}}
Loss: tensor(2.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3813, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8960244648318043, 'agreement': 0.5576923076923077, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.32903225806451614, 'partial': 0.8}, 'recall': {'support': 0.9606557377049181, 'agreement': 0.4084507042253521, 'direct_attack': 0.015384615384615385, 'undercutter_attack': 0.9107142857142857, 'partial': 0.08888888888888889}, 'f1': {'support': 0.9272151898734178, 'agreement': 0.4715447154471545, 'direct_attack': 0.029411764705882356, 'undercutter_attack': 0.48341232227488157, 'partial': 0.15999999999999998}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 56, 'partial': 45}, 'micro_avg': {'precision': 0.6974169741697417, 'recall': 0.6974169741697417, 'f1': 0.6974169741697417, 'support': None}, 'macro_avg': {'precision': 0.5832164727843923, 'recall': 0.476818846383612, 'f1': 0.41431679846026725, 'support': None}, 'weighted_avg': {'precision': 0.7176680604024609, 'recall': 0.6974169741697417, 'f1': 0.6503010377516903, 'support': None}}
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9494949494949495, 'agreement': 0.6170212765957447, 'direct_attack': 0.5, 'undercutter_attack': 0.3333333333333333, 'partial': 0.21428571428571427}, 'recall': {'support': 0.9276315789473685, 'agreement': 0.4084507042253521, 'direct_attack': 0.15384615384615385, 'undercutter_attack': 0.8771929824561403, 'partial': 0.13333333333333333}, 'f1': {'support': 0.9384359400998337, 'agreement': 0.4915254237288136, 'direct_attack': 0.23529411764705882, 'undercutter_attack': 0.4830917874396135, 'partial': 0.1643835616438356}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6955719557195572, 'recall': 0.6955719557195572, 'f1': 0.6955719557195572, 'support': None}, 'macro_avg': {'precision': 0.5228270547419485, 'recall': 0.5000909505616696, 'f1': 0.46254616611183097, 'support': None}, 'weighted_avg': {'precision': 0.7261952627815862, 'recall': 0.6955719557195572, 'f1': 0.6834140971959125, 'support': None}}
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9389830508474576, 'agreement': 0.4666666666666667, 'direct_attack': 0.33076923076923076, 'undercutter_attack': 0.1875, 'partial': 0.18181818181818182}, 'recall': {'support': 0.9111842105263158, 'agreement': 0.5915492957746479, 'direct_attack': 0.6615384615384615, 'undercutter_attack': 0.05263157894736842, 'partial': 0.044444444444444446}, 'f1': {'support': 0.9248747913188647, 'agreement': 0.5217391304347827, 'direct_attack': 0.44102564102564096, 'undercutter_attack': 0.0821917808219178, 'partial': 0.07142857142857142}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6771217712177122, 'recall': 0.6771217712177122, 'f1': 0.6771217712177122, 'support': None}, 'macro_avg': {'precision': 0.4211474260203073, 'recall': 0.45226959824624763, 'f1': 0.40825198300595555, 'support': None}, 'weighted_avg': {'precision': 0.6622758283630602, 'recall': 0.6771217712177122, 'f1': 0.6545595917151404, 'support': None}}
Loss: tensor(1.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9721254355400697, 'agreement': 0.631578947368421, 'direct_attack': 0.3252032520325203, 'undercutter_attack': 0.32653061224489793, 'partial': 0.42857142857142855}, 'recall': {'support': 0.9177631578947368, 'agreement': 0.676056338028169, 'direct_attack': 0.6153846153846154, 'undercutter_attack': 0.2807017543859649, 'partial': 0.06666666666666667}, 'f1': {'support': 0.9441624365482233, 'agreement': 0.6530612244897959, 'direct_attack': 0.425531914893617, 'undercutter_attack': 0.30188679245283023, 'partial': 0.11538461538461539}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7121771217712177, 'recall': 0.7121771217712177, 'f1': 0.7121771217712177, 'support': None}, 'macro_avg': {'precision': 0.5368019351514675, 'recall': 0.5113145064720306, 'f1': 0.4880053967538164, 'support': None}, 'weighted_avg': {'precision': 0.7369085022751407, 'recall': 0.7121771217712177, 'f1': 0.7074763043904787, 'support': None}}
Loss: tensor(2.3031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8895705521472392, 'agreement': 0.5428571428571428, 'direct_attack': 0.3584905660377358, 'undercutter_attack': 0.3191489361702128, 'partial': 0.17857142857142858}, 'recall': {'support': 0.9539473684210527, 'agreement': 0.2676056338028169, 'direct_attack': 0.5846153846153846, 'undercutter_attack': 0.2631578947368421, 'partial': 0.1111111111111111}, 'f1': {'support': 0.9206349206349206, 'agreement': 0.3584905660377358, 'direct_attack': 0.4444444444444444, 'undercutter_attack': 0.28846153846153844, 'partial': 0.13698630136986303}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6771217712177122, 'recall': 0.6771217712177122, 'f1': 0.6771217712177122, 'support': None}, 'macro_avg': {'precision': 0.45772772515675186, 'recall': 0.4360874785374415, 'f1': 0.4298035541897005, 'support': None}, 'weighted_avg': {'precision': 0.6614416889953636, 'recall': 0.6771217712177122, 'f1': 0.6583421147685157, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9116719242902208, 'agreement': 0.5689655172413793, 'direct_attack': 0.40425531914893614, 'undercutter_attack': 0.36507936507936506, 'partial': 0.5}, 'recall': {'support': 0.9506578947368421, 'agreement': 0.4647887323943662, 'direct_attack': 0.5846153846153846, 'undercutter_attack': 0.40350877192982454, 'partial': 0.1111111111111111}, 'f1': {'support': 0.9307568438003221, 'agreement': 0.5116279069767443, 'direct_attack': 0.47798742138364786, 'undercutter_attack': 0.38333333333333336, 'partial': 0.1818181818181818}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7158671586715867, 'recall': 0.7158671586715867, 'f1': 0.7158671586715867, 'support': None}, 'macro_avg': {'precision': 0.5499944251519803, 'recall': 0.5029363789575056, 'f1': 0.4971047374624459, 'support': None}, 'weighted_avg': {'precision': 0.7142637200416414, 'recall': 0.7158671586715867, 'f1': 0.7018019602996349, 'support': None}}
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9166666666666666, 'agreement': 0.5961538461538461, 'direct_attack': 0.3577981651376147, 'undercutter_attack': 0.40350877192982454, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9407894736842105, 'agreement': 0.43661971830985913, 'direct_attack': 0.6, 'undercutter_attack': 0.40350877192982454, 'partial': 0.08888888888888889}, 'f1': {'support': 0.9285714285714286, 'agreement': 0.5040650406504065, 'direct_attack': 0.4482758620689656, 'undercutter_attack': 0.40350877192982454, 'partial': 0.14035087719298245}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7066420664206642, 'recall': 0.7066420664206642, 'f1': 0.7066420664206642, 'support': None}, 'macro_avg': {'precision': 0.5214921566442572, 'recall': 0.4939613705625566, 'f1': 0.48495439608272156, 'support': None}, 'weighted_avg': {'precision': 0.7052591706227578, 'recall': 0.7066420664206642, 'f1': 0.6947012042067531, 'support': None}}
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.988, 'agreement': 0.3776223776223776, 'direct_attack': 0.375, 'undercutter_attack': 0.3372093023255814, 'partial': 0.13333333333333333}, 'recall': {'support': 0.8098360655737705, 'agreement': 0.7714285714285715, 'direct_attack': 0.27692307692307694, 'undercutter_attack': 0.5087719298245614, 'partial': 0.044444444444444446}, 'f1': {'support': 0.8900900900900901, 'agreement': 0.5070422535211268, 'direct_attack': 0.3185840707964602, 'undercutter_attack': 0.4055944055944056, 'partial': 0.06666666666666667}, 'support': {'support': 305, 'agreement': 70, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6457564575645757, 'recall': 0.6457564575645757, 'f1': 0.6457564575645757, 'support': None}, 'macro_avg': {'precision': 0.4422330026562585, 'recall': 0.48228081763888486, 'f1': 0.43759549733374986, 'support': None}, 'weighted_avg': {'precision': 0.6962536838858387, 'recall': 0.6457564575645757, 'f1': 0.6527625109679103, 'support': None}}
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8960244648318043, 'agreement': 0.5217391304347826, 'direct_attack': 0.37142857142857144, 'undercutter_attack': 0.2857142857142857, 'partial': 0.27586206896551724}, 'recall': {'support': 0.9606557377049181, 'agreement': 0.3380281690140845, 'direct_attack': 0.4, 'undercutter_attack': 0.35714285714285715, 'partial': 0.17777777777777778}, 'f1': {'support': 0.9272151898734178, 'agreement': 0.41025641025641024, 'direct_attack': 0.3851851851851852, 'undercutter_attack': 0.3174603174603175, 'partial': 0.2162162162162162}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 56, 'partial': 45}, 'micro_avg': {'precision': 0.6845018450184502, 'recall': 0.6845018450184502, 'f1': 0.6845018450184502, 'support': None}, 'macro_avg': {'precision': 0.47015370427499226, 'recall': 0.44672090832792744, 'f1': 0.45126666379830943, 'support': None}, 'weighted_avg': {'precision': 0.6695342994112091, 'recall': 0.6845018450184502, 'f1': 0.6724601154688968, 'support': None}}
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9174917491749175, 'agreement': 0.3958333333333333, 'direct_attack': 0.3626373626373626, 'undercutter_attack': 0.4, 'partial': 0.5}, 'recall': {'support': 0.9144736842105263, 'agreement': 0.5352112676056338, 'direct_attack': 0.5076923076923077, 'undercutter_attack': 0.2807017543859649, 'partial': 0.13333333333333333}, 'f1': {'support': 0.9159802306425042, 'agreement': 0.45508982035928147, 'direct_attack': 0.4230769230769231, 'undercutter_attack': 0.32989690721649484, 'partial': 0.2105263157894737}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6845018450184502, 'recall': 0.6845018450184502, 'f1': 0.6845018450184502, 'support': None}, 'macro_avg': {'precision': 0.5151924890291226, 'recall': 0.4742824694455532, 'f1': 0.46691403941693554, 'support': None}, 'weighted_avg': {'precision': 0.6935296807883213, 'recall': 0.6845018450184502, 'f1': 0.6762863012595882, 'support': None}}
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8773006134969326, 'agreement': 0.3877551020408163, 'direct_attack': 0.40217391304347827, 'undercutter_attack': 0.3333333333333333, 'partial': 0.3888888888888889}, 'recall': {'support': 0.9407894736842105, 'agreement': 0.2676056338028169, 'direct_attack': 0.5692307692307692, 'undercutter_attack': 0.3333333333333333, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9079365079365079, 'agreement': 0.31666666666666665, 'direct_attack': 0.4713375796178344, 'undercutter_attack': 0.3333333333333333, 'partial': 0.22222222222222218}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6789667896678967, 'recall': 0.6789667896678967, 'f1': 0.6789667896678967, 'support': None}, 'macro_avg': {'precision': 0.47789037016068986, 'recall': 0.4533029531213371, 'f1': 0.45029926195531295, 'support': None}, 'weighted_avg': {'precision': 0.6584341385531208, 'recall': 0.6789667896678967, 'f1': 0.6607619454265516, 'support': None}}
Loss: tensor(9.6490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9344262295081968, 'agreement': 0.5789473684210527, 'direct_attack': 0.38095238095238093, 'undercutter_attack': 0.35365853658536583, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9375, 'agreement': 0.4647887323943662, 'direct_attack': 0.36923076923076925, 'undercutter_attack': 0.5087719298245614, 'partial': 0.2222222222222222}, 'f1': {'support': 0.935960591133005, 'agreement': 0.515625, 'direct_attack': 0.375, 'undercutter_attack': 0.4172661870503597, 'partial': 0.25}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7029520295202952, 'recall': 0.7029520295202952, 'f1': 0.7029520295202952, 'support': None}, 'macro_avg': {'precision': 0.5067397602362563, 'recall': 0.5005027307343839, 'f1': 0.49877035563667294, 'support': None}, 'weighted_avg': {'precision': 0.7065469024590405, 'recall': 0.7029520295202952, 'f1': 0.7021228180190112, 'support': None}}
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(12.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9466666666666667, 'agreement': 0.5263157894736842, 'direct_attack': 0.3372093023255814, 'undercutter_attack': 0.3584905660377358, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9311475409836065, 'agreement': 0.5633802816901409, 'direct_attack': 0.4461538461538462, 'undercutter_attack': 0.3392857142857143, 'partial': 0.2}, 'f1': {'support': 0.9388429752066114, 'agreement': 0.54421768707483, 'direct_attack': 0.3841059602649006, 'undercutter_attack': 0.34862385321100914, 'partial': 0.25}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 56, 'partial': 45}, 'micro_avg': {'precision': 0.7029520295202952, 'recall': 0.7029520295202952, 'f1': 0.7029520295202952, 'support': None}, 'macro_avg': {'precision': 0.5004031315674002, 'recall': 0.4959934766226616, 'f1': 0.4931580951514702, 'support': None}, 'weighted_avg': {'precision': 0.7068188758952785, 'recall': 0.7029520295202952, 'f1': 0.7024472074121116, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9563492063492064, 'agreement': 0.5294117647058824, 'direct_attack': 0.22549019607843138, 'undercutter_attack': 0.36470588235294116, 'partial': 0.21153846153846154}, 'recall': {'support': 0.7901639344262295, 'agreement': 0.38028169014084506, 'direct_attack': 0.35384615384615387, 'undercutter_attack': 0.543859649122807, 'partial': 0.25}, 'f1': {'support': 0.8653500897666068, 'agreement': 0.4426229508196721, 'direct_attack': 0.2754491017964072, 'undercutter_attack': 0.4366197183098592, 'partial': 0.22916666666666666}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 44}, 'micro_avg': {'precision': 0.6143911439114391, 'recall': 0.6143911439114391, 'f1': 0.6143911439114391, 'support': None}, 'macro_avg': {'precision': 0.45749910220498463, 'recall': 0.4636302855072071, 'f1': 0.4498417054718423, 'support': None}, 'weighted_avg': {'precision': 0.6900877003275528, 'recall': 0.6143911439114391, 'f1': 0.6424960438759659, 'support': None}}
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9726027397260274, 'agreement': 0.5588235294117647, 'direct_attack': 0.37894736842105264, 'undercutter_attack': 0.3870967741935484, 'partial': 0.36}, 'recall': {'support': 0.9342105263157895, 'agreement': 0.5352112676056338, 'direct_attack': 0.5538461538461539, 'undercutter_attack': 0.42105263157894735, 'partial': 0.2}, 'f1': {'support': 0.9530201342281879, 'agreement': 0.5467625899280575, 'direct_attack': 0.45, 'undercutter_attack': 0.4033613445378151, 'partial': 0.2571428571428571}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7214022140221402, 'recall': 0.7214022140221402, 'f1': 0.7214022140221401, 'support': None}, 'macro_avg': {'precision': 0.5314940823504786, 'recall': 0.5288641158693049, 'f1': 0.5220573851673835, 'support': None}, 'weighted_avg': {'precision': 0.7347671559803475, 'recall': 0.7214022140221402, 'f1': 0.7238953688198251, 'support': None}}
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9812734082397003, 'agreement': 0.4824561403508772, 'direct_attack': 0.34782608695652173, 'undercutter_attack': 0.3611111111111111, 'partial': 0.5714285714285714}, 'recall': {'support': 0.8590163934426229, 'agreement': 0.7857142857142857, 'direct_attack': 0.24615384615384617, 'undercutter_attack': 0.6842105263157895, 'partial': 0.08888888888888889}, 'f1': {'support': 0.916083916083916, 'agreement': 0.5978260869565217, 'direct_attack': 0.2882882882882883, 'undercutter_attack': 0.4727272727272728, 'partial': 0.15384615384615385}, 'support': {'support': 305, 'agreement': 70, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6937269372693727, 'recall': 0.6937269372693727, 'f1': 0.6937269372693727, 'support': None}, 'macro_avg': {'precision': 0.5488190636173563, 'recall': 0.5327967881030866, 'f1': 0.4857543435804305, 'support': None}, 'weighted_avg': {'precision': 0.7416358561576808, 'recall': 0.6937269372693727, 'f1': 0.6897798721398913, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.9855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9773584905660377, 'agreement': 0.5595238095238095, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.3178294573643411, 'partial': 0.3181818181818182}, 'recall': {'support': 0.8519736842105263, 'agreement': 0.6619718309859155, 'direct_attack': 0.2153846153846154, 'undercutter_attack': 0.7192982456140351, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9103690685413006, 'agreement': 0.6064516129032258, 'direct_attack': 0.2616822429906542, 'undercutter_attack': 0.4408602150537634, 'partial': 0.208955223880597}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6789667896678967, 'recall': 0.6789667896678967, 'f1': 0.6789667896678967, 'support': None}, 'macro_avg': {'precision': 0.501245381793868, 'recall': 0.5208367863501295, 'f1': 0.4856636726739082, 'support': None}, 'weighted_avg': {'precision': 0.7212994449499667, 'recall': 0.6789667896678967, 'f1': 0.6851505986711593, 'support': None}}
Loss: tensor(0.4541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(10.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9435215946843853, 'agreement': 0.5789473684210527, 'direct_attack': 0.4457831325301205, 'undercutter_attack': 0.38235294117647056, 'partial': 0.24242424242424243}, 'recall': {'support': 0.9311475409836065, 'agreement': 0.4647887323943662, 'direct_attack': 0.5692307692307692, 'undercutter_attack': 0.45614035087719296, 'partial': 0.18181818181818182}, 'f1': {'support': 0.9372937293729373, 'agreement': 0.515625, 'direct_attack': 0.5, 'undercutter_attack': 0.4159999999999999, 'partial': 0.2077922077922078}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 44}, 'micro_avg': {'precision': 0.7158671586715867, 'recall': 0.7158671586715867, 'f1': 0.7158671586715867, 'support': None}, 'macro_avg': {'precision': 0.5186058558472542, 'recall': 0.5206251150608233, 'f1': 0.515342187433029, 'support': None}, 'weighted_avg': {'precision': 0.7201402905254901, 'recall': 0.7158671586715867, 'f1': 0.7155697778627362, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9273927392739274, 'agreement': 0.5, 'direct_attack': 0.3763440860215054, 'undercutter_attack': 0.38596491228070173, 'partial': 0.30434782608695654}, 'recall': {'support': 0.9243421052631579, 'agreement': 0.4647887323943662, 'direct_attack': 0.5384615384615384, 'undercutter_attack': 0.38596491228070173, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9258649093904449, 'agreement': 0.4817518248175182, 'direct_attack': 0.44303797468354433, 'undercutter_attack': 0.38596491228070173, 'partial': 0.2058823529411765}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6974169741697417, 'recall': 0.6974169741697417, 'f1': 0.6974169741697417, 'support': None}, 'macro_avg': {'precision': 0.49880991273261815, 'recall': 0.4938225687910639, 'f1': 0.4885003948226771, 'support': None}, 'weighted_avg': {'precision': 0.696652048901448, 'recall': 0.6974169741697417, 'f1': 0.693227834416093, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9278688524590164, 'agreement': 0.47761194029850745, 'direct_attack': 0.37037037037037035, 'undercutter_attack': 0.36923076923076925, 'partial': 0.2916666666666667}, 'recall': {'support': 0.930921052631579, 'agreement': 0.4507042253521127, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.42105263157894735, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9293924466338259, 'agreement': 0.463768115942029, 'direct_attack': 0.4109589041095891, 'undercutter_attack': 0.39344262295081966, 'partial': 0.2028985507246377}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6937269372693727, 'recall': 0.6937269372693727, 'f1': 0.6937269372693727, 'support': None}, 'macro_avg': {'precision': 0.487349719805066, 'recall': 0.4839543853313312, 'f1': 0.48009212807218027, 'support': None}, 'weighted_avg': {'precision': 0.6904572081715182, 'recall': 0.6937269372693727, 'f1': 0.689542127428959, 'support': None}}
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9276315789473685, 'agreement': 0.47761194029850745, 'direct_attack': 0.375, 'undercutter_attack': 0.3787878787878788, 'partial': 0.28}, 'recall': {'support': 0.9276315789473685, 'agreement': 0.4507042253521127, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9276315789473685, 'agreement': 0.463768115942029, 'direct_attack': 0.41379310344827586, 'undercutter_attack': 0.4065040650406504, 'partial': 0.2}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6937269372693727, 'recall': 0.6937269372693727, 'f1': 0.6937269372693727, 'support': None}, 'macro_avg': {'precision': 0.4878062796067509, 'recall': 0.48680526252431366, 'f1': 0.48233937267566473, 'support': None}, 'weighted_avg': {'precision': 0.6909157875500057, 'recall': 0.6937269372693727, 'f1': 0.690027342552286, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9276315789473685, 'agreement': 0.47058823529411764, 'direct_attack': 0.38961038961038963, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9276315789473685, 'agreement': 0.4507042253521127, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9276315789473685, 'agreement': 0.460431654676259, 'direct_attack': 0.4225352112676056, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6937269372693727, 'recall': 0.6937269372693727, 'f1': 0.6937269372693727, 'support': None}, 'macro_avg': {'precision': 0.48709545253508113, 'recall': 0.48680526252431366, 'f1': 0.4821196889782466, 'support': None}, 'weighted_avg': {'precision': 0.690576240559961, 'recall': 0.6937269372693727, 'f1': 0.6899546793623779, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9278688524590164, 'agreement': 0.4626865671641791, 'direct_attack': 0.38961038961038963, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9278688524590164, 'agreement': 0.44285714285714284, 'direct_attack': 0.46153846153846156, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9278688524590164, 'agreement': 0.45255474452554745, 'direct_attack': 0.4225352112676056, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 305, 'agreement': 70, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6937269372693727, 'recall': 0.6937269372693727, 'f1': 0.6937269372693727, 'support': None}, 'macro_avg': {'precision': 0.4855625736114229, 'recall': 0.48528330072764925, 'f1': 0.48059176165043394, 'support': None}, 'weighted_avg': {'precision': 0.6905325043895, 'recall': 0.6937269372693727, 'f1': 0.6899328797955401, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9306930693069307, 'agreement': 0.47058823529411764, 'direct_attack': 0.3974358974358974, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9276315789473685, 'agreement': 0.4507042253521127, 'direct_attack': 0.47692307692307695, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9291598023064251, 'agreement': 0.460431654676259, 'direct_attack': 0.4335664335664336, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6955719557195572, 'recall': 0.6955719557195572, 'f1': 0.6955719557195572, 'support': None}, 'macro_avg': {'precision': 0.4892728521720951, 'recall': 0.48988218560123675, 'f1': 0.48463157810982355, 'support': None}, 'weighted_avg': {'precision': 0.6932318698550992, 'recall': 0.6955719557195572, 'f1': 0.6921347704151031, 'support': None}}
Loss: tensor(8.8809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9011e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2366e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6605e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9306930693069307, 'agreement': 0.47761194029850745, 'direct_attack': 0.4050632911392405, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9276315789473685, 'agreement': 0.4507042253521127, 'direct_attack': 0.49230769230769234, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9291598023064251, 'agreement': 0.463768115942029, 'direct_attack': 0.4444444444444445, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6974169741697417, 'recall': 0.6974169741697417, 'f1': 0.6974169741697417, 'support': None}, 'macro_avg': {'precision': 0.49220307191364154, 'recall': 0.49295910867815984, 'f1': 0.48747447253857973, 'support': None}, 'weighted_avg': {'precision': 0.6950666736300605, 'recall': 0.6974169741697417, 'f1': 0.6938763930293842, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6286e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3569e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9306930693069307, 'agreement': 0.48484848484848486, 'direct_attack': 0.4, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9276315789473685, 'agreement': 0.4507042253521127, 'direct_attack': 0.49230769230769234, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9291598023064251, 'agreement': 0.46715328467153283, 'direct_attack': 0.44137931034482764, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6974169741697417, 'recall': 0.6974169741697417, 'f1': 0.6974169741697417, 'support': None}, 'macro_avg': {'precision': 0.4926377225957889, 'recall': 0.49295910867815984, 'f1': 0.4875384794645571, 'support': None}, 'weighted_avg': {'precision': 0.695407413000905, 'recall': 0.6974169741697417, 'f1': 0.6939522477587562, 'support': None}}
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7800e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3511e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.930921052631579, 'agreement': 0.49230769230769234, 'direct_attack': 0.4, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.930921052631579, 'agreement': 0.4507042253521127, 'direct_attack': 0.49230769230769234, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.930921052631579, 'agreement': 0.47058823529411764, 'direct_attack': 0.44137931034482764, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 304, 'agreement': 71, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6992619926199262, 'recall': 0.6992619926199262, 'f1': 0.6992619926199262, 'support': None}, 'macro_avg': {'precision': 0.4941751607525601, 'recall': 0.4936170034150019, 'f1': 0.48857771965410485, 'support': None}, 'weighted_avg': {'precision': 0.6965124142191648, 'recall': 0.6992619926199262, 'f1': 0.6953900735761922, 'support': None}}
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9665e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0852e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5983e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7497e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9342105263157895, 'agreement': 0.49230769230769234, 'direct_attack': 0.4, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9311475409836065, 'agreement': 0.45714285714285713, 'direct_attack': 0.49230769230769234, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9326765188834155, 'agreement': 0.4740740740740741, 'direct_attack': 0.44137931034482764, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 305, 'agreement': 70, 'direct_attack': 65, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7011070110701108, 'recall': 0.7011070110701108, 'f1': 0.7011070110701108, 'support': None}, 'macro_avg': {'precision': 0.49483305548940226, 'recall': 0.49495002744355626, 'f1': 0.4896259806604634, 'support': None}, 'weighted_avg': {'precision': 0.699172751551283, 'recall': 0.7011070110701108, 'f1': 0.6976774513229533, 'support': None}}
Loss: tensor(4.0877e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6605e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9330e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6294e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9311475409836065, 'agreement': 0.49230769230769234, 'direct_attack': 0.3924050632911392, 'undercutter_attack': 0.36764705882352944, 'partial': 0.28}, 'recall': {'support': 0.9311475409836065, 'agreement': 0.4507042253521127, 'direct_attack': 0.484375, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9311475409836065, 'agreement': 0.47058823529411764, 'direct_attack': 0.43356643356643354, 'undercutter_attack': 0.4, 'partial': 0.2}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 64, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.6992619926199262, 'recall': 0.6992619926199262, 'f1': 0.6992619926199262, 'support': None}, 'macro_avg': {'precision': 0.49270147108119355, 'recall': 0.49207576262386893, 'f1': 0.48706044196883164, 'support': None}, 'weighted_avg': {'precision': 0.6967226061945023, 'recall': 0.6992619926199262, 'f1': 0.6954981853397308, 'support': None}}
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2701e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2047e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6294e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5083e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9322e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5132e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7554e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0279e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3888e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0893e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5734e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7849e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6343e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5426e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1744e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4780e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4510e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4215e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3913e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6335e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9379e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8765e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9311475409836065, 'agreement': 0.5076923076923077, 'direct_attack': 0.4, 'undercutter_attack': 0.36764705882352944, 'partial': 0.2916666666666667}, 'recall': {'support': 0.9311475409836065, 'agreement': 0.4647887323943662, 'direct_attack': 0.5, 'undercutter_attack': 0.43859649122807015, 'partial': 0.15555555555555556}, 'f1': {'support': 0.9311475409836065, 'agreement': 0.48529411764705876, 'direct_attack': 0.4444444444444445, 'undercutter_attack': 0.4, 'partial': 0.2028985507246377}, 'support': {'support': 305, 'agreement': 71, 'direct_attack': 64, 'undercutter_attack': 57, 'partial': 45}, 'micro_avg': {'precision': 0.7029520295202952, 'recall': 0.7029520295202952, 'f1': 0.7029520295202952, 'support': None}, 'macro_avg': {'precision': 0.499630714833222, 'recall': 0.4980176640323196, 'f1': 0.49275693075994953, 'support': None}, 'weighted_avg': {'precision': 0.7006033878212086, 'recall': 0.7029520295202952, 'f1': 0.6989497446125357, 'support': None}}


		-------------RUN 5-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.5394321766561514, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.07804878048780488, 'partial': 0.0}, 'recall': {'support': 0.5738255033557047, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.25396825396825395, 'partial': 0.0}, 'f1': {'support': 0.5560975609756097, 'agreement': 0.0, 'direct_attack': 0.0, 'undercutter_attack': 0.11940298507462686, 'partial': 0.0}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 63, 'partial': 37}, 'micro_avg': {'precision': 0.35823754789272033, 'recall': 0.35823754789272033, 'f1': 0.35823754789272033, 'support': None}, 'macro_avg': {'precision': 0.12349619142879124, 'recall': 0.16555875146479174, 'f1': 0.1351001092100473, 'support': None}, 'weighted_avg': {'precision': 0.31737138278594795, 'recall': 0.35823754789272033, 'f1': 0.33187636251040836, 'support': None}}
Loss: tensor(0.7071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8553, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.0, 'agreement': 0.1475095785440613, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.0, 'agreement': 1.0, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.0, 'agreement': 0.2570951585976628, 'direct_attack': 0.0, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.1475095785440613, 'recall': 0.1475095785440613, 'f1': 0.1475095785440613, 'support': None}, 'macro_avg': {'precision': 0.02950191570881226, 'recall': 0.2, 'f1': 0.05141903171953256, 'support': None}, 'weighted_avg': {'precision': 0.02175907576224659, 'recall': 0.1475095785440613, 'f1': 0.03792399849045984, 'support': None}}
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3714, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8585526315789473, 'agreement': 0.75, 'direct_attack': 0.0, 'undercutter_attack': 0.43617021276595747, 'partial': 0.15714285714285714}, 'recall': {'support': 0.8787878787878788, 'agreement': 0.5064935064935064, 'direct_attack': 0.0, 'undercutter_attack': 0.640625, 'partial': 0.2972972972972973}, 'f1': {'support': 0.8685524126455907, 'agreement': 0.6046511627906975, 'direct_attack': 0.0, 'undercutter_attack': 0.5189873417721519, 'partial': 0.205607476635514}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6743295019157088, 'recall': 0.6743295019157088, 'f1': 0.6743295019157088, 'support': None}, 'macro_avg': {'precision': 0.44037314029755237, 'recall': 0.46464073651573656, 'f1': 0.4395596787687908, 'support': None}, 'weighted_avg': {'precision': 0.6637343120886099, 'recall': 0.6743295019157088, 'f1': 0.6615725528727123, 'support': None}}
Loss: tensor(2.5949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9611, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8730650154798761, 'agreement': 0.7777777777777778, 'direct_attack': 0.07142857142857142, 'undercutter_attack': 0.43820224719101125, 'partial': 0.14285714285714285}, 'recall': {'support': 0.9494949494949495, 'agreement': 0.5454545454545454, 'direct_attack': 0.0425531914893617, 'undercutter_attack': 0.609375, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9096774193548388, 'agreement': 0.6412213740458015, 'direct_attack': 0.05333333333333332, 'undercutter_attack': 0.5098039215686275, 'partial': 0.12307692307692308}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7068965517241379, 'recall': 0.7068965517241379, 'f1': 0.7068965517241379, 'support': None}, 'macro_avg': {'precision': 0.46066615094687585, 'recall': 0.45099715890939296, 'f1': 0.4474225942759048, 'support': None}, 'weighted_avg': {'precision': 0.6817567039262337, 'recall': 0.7068965517241379, 'f1': 0.6881919600590399, 'support': None}}
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8392857142857143, 'agreement': 0.75, 'direct_attack': 0.0, 'undercutter_attack': 0.42727272727272725, 'partial': 0.14705882352941177}, 'recall': {'support': 0.9494949494949495, 'agreement': 0.3116883116883117, 'direct_attack': 0.0, 'undercutter_attack': 0.734375, 'partial': 0.13513513513513514}, 'f1': {'support': 0.8909952606635071, 'agreement': 0.4403669724770642, 'direct_attack': 0.0, 'undercutter_attack': 0.5402298850574713, 'partial': 0.14084507042253522}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.685823754789272, 'recall': 0.685823754789272, 'f1': 0.685823754789272, 'support': None}, 'macro_avg': {'precision': 0.43272345301757065, 'recall': 0.42613867926367927, 'f1': 0.40248743772411555, 'support': None}, 'weighted_avg': {'precision': 0.6509664524116857, 'recall': 0.685823754789272, 'f1': 0.6481222788258764, 'support': None}}
Loss: tensor(1.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(13.6620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8862876254180602, 'agreement': 0.6153846153846154, 'direct_attack': 0.16326530612244897, 'undercutter_attack': 0.3148148148148148, 'partial': 0.19047619047619047}, 'recall': {'support': 0.8922558922558923, 'agreement': 0.6233766233766234, 'direct_attack': 0.1702127659574468, 'undercutter_attack': 0.265625, 'partial': 0.21621621621621623}, 'f1': {'support': 0.8892617449664428, 'agreement': 0.6193548387096774, 'direct_attack': 0.16666666666666666, 'undercutter_attack': 0.28813559322033894, 'partial': 0.20253164556962025}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6628352490421456, 'recall': 0.6628352490421456, 'f1': 0.6628352490421456, 'support': None}, 'macro_avg': {'precision': 0.434045710443226, 'recall': 0.43353729956123577, 'f1': 0.43319009782654927, 'support': None}, 'weighted_avg': {'precision': 0.6618415262783555, 'recall': 0.6628352490421456, 'f1': 0.6620090862474898, 'support': None}}
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8507462686567164, 'agreement': 0.6, 'direct_attack': 0.20408163265306123, 'undercutter_attack': 0.41379310344827586, 'partial': 0.2}, 'recall': {'support': 0.9563758389261745, 'agreement': 0.5844155844155844, 'direct_attack': 0.21739130434782608, 'undercutter_attack': 0.375, 'partial': 0.02702702702702703}, 'f1': {'support': 0.9004739336492892, 'agreement': 0.5921052631578947, 'direct_attack': 0.2105263157894737, 'undercutter_attack': 0.3934426229508197, 'partial': 0.047619047619047616}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6992337164750958, 'recall': 0.6992337164750958, 'f1': 0.6992337164750958, 'support': None}, 'macro_avg': {'precision': 0.4537242009516108, 'recall': 0.4320419509433225, 'f1': 0.428833436633305, 'support': None}, 'weighted_avg': {'precision': 0.6570745244874175, 'recall': 0.6992337164750958, 'f1': 0.6715704610109561, 'support': None}}
Loss: tensor(0.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8482142857142857, 'agreement': 0.7777777777777778, 'direct_attack': 0.27380952380952384, 'undercutter_attack': 0.4897959183673469, 'partial': 0.375}, 'recall': {'support': 0.9563758389261745, 'agreement': 0.45454545454545453, 'direct_attack': 0.5, 'undercutter_attack': 0.375, 'partial': 0.08108108108108109}, 'f1': {'support': 0.8990536277602523, 'agreement': 0.5737704918032788, 'direct_attack': 0.35384615384615387, 'undercutter_attack': 0.42477876106194684, 'partial': 0.13333333333333333}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7088122605363985, 'recall': 0.7088122605363985, 'f1': 0.7088122605363985, 'support': None}, 'macro_avg': {'precision': 0.5529195011337869, 'recall': 0.4734004749105421, 'f1': 0.476956473560993, 'support': None}, 'weighted_avg': {'precision': 0.7097201588170389, 'recall': 0.7088122605363985, 'f1': 0.6906023104590588, 'support': None}}
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8367346938775511, 'agreement': 0.8095238095238095, 'direct_attack': 0.21875, 'undercutter_attack': 0.44565217391304346, 'partial': 0.23076923076923078}, 'recall': {'support': 0.9663299663299664, 'agreement': 0.44155844155844154, 'direct_attack': 0.14893617021276595, 'undercutter_attack': 0.640625, 'partial': 0.08108108108108109}, 'f1': {'support': 0.896875, 'agreement': 0.5714285714285715, 'direct_attack': 0.1772151898734177, 'undercutter_attack': 0.5256410256410257, 'partial': 0.12}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7126436781609196, 'recall': 0.7126436781609196, 'f1': 0.7126436781609196, 'support': None}, 'macro_avg': {'precision': 0.508285981616727, 'recall': 0.455706131836451, 'f1': 0.458231957388603, 'support': None}, 'weighted_avg': {'precision': 0.6861781380916903, 'recall': 0.7126436781609196, 'f1': 0.6834904493583837, 'support': None}}
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8313953488372093, 'agreement': 0.84375, 'direct_attack': 0.25925925925925924, 'undercutter_attack': 0.4424778761061947, 'partial': 0.16666666666666666}, 'recall': {'support': 0.959731543624161, 'agreement': 0.35064935064935066, 'direct_attack': 0.14893617021276595, 'undercutter_attack': 0.78125, 'partial': 0.027777777777777776}, 'f1': {'support': 0.8909657320872274, 'agreement': 0.4954128440366973, 'direct_attack': 0.18918918918918917, 'undercutter_attack': 0.5649717514124294, 'partial': 0.04761904761904762}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 36}, 'micro_avg': {'precision': 0.710727969348659, 'recall': 0.710727969348659, 'f1': 0.710727969348659, 'support': None}, 'macro_avg': {'precision': 0.508709830173866, 'recall': 0.45366896845281107, 'f1': 0.4376317128689181, 'support': None}, 'weighted_avg': {'precision': 0.688176883543046, 'recall': 0.710727969348659, 'f1': 0.6713006644624377, 'support': None}}
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8362573099415205, 'agreement': 0.7575757575757576, 'direct_attack': 0.21052631578947367, 'undercutter_attack': 0.46875, 'partial': 0.28125}, 'recall': {'support': 0.959731543624161, 'agreement': 0.3246753246753247, 'direct_attack': 0.08695652173913043, 'undercutter_attack': 0.703125, 'partial': 0.24324324324324326}, 'f1': {'support': 0.8937499999999999, 'agreement': 0.45454545454545453, 'direct_attack': 0.12307692307692308, 'undercutter_attack': 0.5625, 'partial': 0.26086956521739135}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7068965517241379, 'recall': 0.7068965517241379, 'f1': 0.7068965517241379, 'support': None}, 'macro_avg': {'precision': 0.5108718766613503, 'recall': 0.4635463266563719, 'f1': 0.45894838856795384, 'support': None}, 'weighted_avg': {'precision': 0.6851120157513836, 'recall': 0.7068965517241379, 'f1': 0.6755770352003485, 'support': None}}
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.844776119402985, 'agreement': 0.7090909090909091, 'direct_attack': 0.19230769230769232, 'undercutter_attack': 0.4142857142857143, 'partial': 0.3}, 'recall': {'support': 0.9496644295302014, 'agreement': 0.5131578947368421, 'direct_attack': 0.2127659574468085, 'undercutter_attack': 0.453125, 'partial': 0.08108108108108109}, 'f1': {'support': 0.8941548183254344, 'agreement': 0.5954198473282444, 'direct_attack': 0.20202020202020202, 'undercutter_attack': 0.43283582089552236, 'partial': 0.12765957446808512}, 'support': {'support': 298, 'agreement': 76, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6973180076628352, 'recall': 0.6973180076628352, 'f1': 0.6973180076628352, 'support': None}, 'macro_avg': {'precision': 0.49209208701746016, 'recall': 0.44195887255898664, 'f1': 0.4504180526074976, 'support': None}, 'weighted_avg': {'precision': 0.6748791952600497, 'recall': 0.6973180076628352, 'f1': 0.6774518975967588, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8402366863905325, 'agreement': 0.75, 'direct_attack': 0.23255813953488372, 'undercutter_attack': 0.41304347826086957, 'partial': 0.375}, 'recall': {'support': 0.9530201342281879, 'agreement': 0.35064935064935066, 'direct_attack': 0.43478260869565216, 'undercutter_attack': 0.296875, 'partial': 0.16216216216216217}, 'f1': {'support': 0.8930817610062892, 'agreement': 0.4778761061946903, 'direct_attack': 0.30303030303030304, 'undercutter_attack': 0.3454545454545454, 'partial': 0.22641509433962265}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6819923371647509, 'recall': 0.6819923371647509, 'f1': 0.6819923371647509, 'support': None}, 'macro_avg': {'precision': 0.5221676608372572, 'recall': 0.4394978511470706, 'f1': 0.44917156200509006, 'support': None}, 'weighted_avg': {'precision': 0.6880229685281206, 'recall': 0.6819923371647509, 'f1': 0.6654418932871958, 'support': None}}
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8431952662721893, 'agreement': 0.9166666666666666, 'direct_attack': 0.19444444444444445, 'undercutter_attack': 0.4230769230769231, 'partial': 0.25}, 'recall': {'support': 0.9595959595959596, 'agreement': 0.2857142857142857, 'direct_attack': 0.14893617021276595, 'undercutter_attack': 0.34375, 'partial': 0.4864864864864865}, 'f1': {'support': 0.8976377952755906, 'agreement': 0.4356435643564356, 'direct_attack': 0.16867469879518074, 'undercutter_attack': 0.3793103448275862, 'partial': 0.3302752293577982}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6781609195402298, 'recall': 0.6781609195402298, 'f1': 0.6781609195402298, 'support': None}, 'macro_avg': {'precision': 0.5254766600920446, 'recall': 0.44489658040189956, 'f1': 0.4423083265225182, 'support': None}, 'weighted_avg': {'precision': 0.7020654011149149, 'recall': 0.6781609195402298, 'f1': 0.6600895326645088, 'support': None}}
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8411764705882353, 'agreement': 0.75, 'direct_attack': 0.21311475409836064, 'undercutter_attack': 0.3783783783783784, 'partial': 0.36363636363636365}, 'recall': {'support': 0.959731543624161, 'agreement': 0.35064935064935066, 'direct_attack': 0.2765957446808511, 'undercutter_attack': 0.4444444444444444, 'partial': 0.10810810810810811}, 'f1': {'support': 0.896551724137931, 'agreement': 0.4778761061946903, 'direct_attack': 0.24074074074074073, 'undercutter_attack': 0.4087591240875913, 'partial': 0.16666666666666669}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 63, 'partial': 37}, 'micro_avg': {'precision': 0.685823754789272, 'recall': 0.685823754789272, 'f1': 0.685823754789272, 'support': None}, 'macro_avg': {'precision': 0.5092611933402676, 'recall': 0.42790583830138296, 'f1': 0.43811887236552394, 'support': None}, 'weighted_avg': {'precision': 0.6814738792534488, 'recall': 0.685823754789272, 'f1': 0.6651382763775754, 'support': None}}
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8623853211009175, 'agreement': 0.7111111111111111, 'direct_attack': 0.17073170731707318, 'undercutter_attack': 0.4358974358974359, 'partial': 0.3225806451612903}, 'recall': {'support': 0.9463087248322147, 'agreement': 0.4155844155844156, 'direct_attack': 0.15217391304347827, 'undercutter_attack': 0.53125, 'partial': 0.2702702702702703}, 'f1': {'support': 0.9024, 'agreement': 0.5245901639344263, 'direct_attack': 0.16091954022988506, 'undercutter_attack': 0.47887323943661975, 'partial': 0.2941176470588235}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6992337164750958, 'recall': 0.6992337164750958, 'f1': 0.6992337164750958, 'support': None}, 'macro_avg': {'precision': 0.5005412441175656, 'recall': 0.46311746474607574, 'f1': 0.4721801181319509, 'support': None}, 'weighted_avg': {'precision': 0.6885688880241725, 'recall': 0.6992337164750958, 'f1': 0.6862857887713518, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8252148997134671, 'agreement': 0.8888888888888888, 'direct_attack': 0.24390243902439024, 'undercutter_attack': 0.41509433962264153, 'partial': 0.25}, 'recall': {'support': 0.9664429530201343, 'agreement': 0.21052631578947367, 'direct_attack': 0.425531914893617, 'undercutter_attack': 0.34375, 'partial': 0.13513513513513514}, 'f1': {'support': 0.8902627511591963, 'agreement': 0.3404255319148936, 'direct_attack': 0.310077519379845, 'undercutter_attack': 0.37606837606837606, 'partial': 0.1754385964912281}, 'support': {'support': 298, 'agreement': 76, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6724137931034483, 'recall': 0.6724137931034483, 'f1': 0.6724137931034483, 'support': None}, 'macro_avg': {'precision': 0.5246201134498776, 'recall': 0.41627726376767205, 'f1': 0.4184545550027078, 'support': None}, 'weighted_avg': {'precision': 0.6910901303451421, 'recall': 0.6724137931034483, 'f1': 0.6442603214949744, 'support': None}}
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8706624605678234, 'agreement': 0.6666666666666666, 'direct_attack': 0.16666666666666666, 'undercutter_attack': 0.44680851063829785, 'partial': 0.2962962962962963}, 'recall': {'support': 0.9292929292929293, 'agreement': 0.4155844155844156, 'direct_attack': 0.1276595744680851, 'undercutter_attack': 0.65625, 'partial': 0.21621621621621623}, 'f1': {'support': 0.8990228013029316, 'agreement': 0.512, 'direct_attack': 0.14457831325301204, 'undercutter_attack': 0.5316455696202532, 'partial': 0.25}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6973180076628352, 'recall': 0.6973180076628352, 'f1': 0.6973180076628352, 'support': None}, 'macro_avg': {'precision': 0.48942012016715014, 'recall': 0.4690006271123292, 'f1': 0.46744933683523937, 'support': None}, 'weighted_avg': {'precision': 0.6845059867799314, 'recall': 0.6973180076628352, 'f1': 0.6829583700489625, 'support': None}}
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8362573099415205, 'agreement': 0.8, 'direct_attack': 0.2033898305084746, 'undercutter_attack': 0.43243243243243246, 'partial': 0.25}, 'recall': {'support': 0.959731543624161, 'agreement': 0.3684210526315789, 'direct_attack': 0.2553191489361702, 'undercutter_attack': 0.5, 'partial': 0.08108108108108109}, 'f1': {'support': 0.8937499999999999, 'agreement': 0.5045045045045045, 'direct_attack': 0.22641509433962265, 'undercutter_attack': 0.463768115942029, 'partial': 0.12244897959183675}, 'support': {'support': 298, 'agreement': 76, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.6915708812260536, 'recall': 0.6915708812260536, 'f1': 0.6915708812260536, 'support': None}, 'macro_avg': {'precision': 0.5044159145764855, 'recall': 0.43291056525459826, 'f1': 0.44217733887559857, 'support': None}, 'weighted_avg': {'precision': 0.6829304139313163, 'recall': 0.6915708812260536, 'f1': 0.6696036847538169, 'support': None}}
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8567073170731707, 'agreement': 0.7708333333333334, 'direct_attack': 0.234375, 'undercutter_attack': 0.410958904109589, 'partial': 0.5555555555555556}, 'recall': {'support': 0.9461279461279462, 'agreement': 0.4805194805194805, 'direct_attack': 0.3191489361702128, 'undercutter_attack': 0.46875, 'partial': 0.13513513513513514}, 'f1': {'support': 0.8992, 'agreement': 0.592, 'direct_attack': 0.2702702702702703, 'undercutter_attack': 0.43795620437956206, 'partial': 0.2173913043478261}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7049808429118773, 'recall': 0.7049808429118773, 'f1': 0.7049808429118773, 'support': None}, 'macro_avg': {'precision': 0.5656860220143297, 'recall': 0.4699362995905549, 'f1': 0.48336355579953166, 'support': None}, 'weighted_avg': {'precision': 0.7120091767355702, 'recall': 0.7049808429118773, 'f1': 0.6923788851415025, 'support': None}}
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8461538461538461, 'agreement': 0.8780487804878049, 'direct_attack': 0.24074074074074073, 'undercutter_attack': 0.3924050632911392, 'partial': 0.3}, 'recall': {'support': 0.9629629629629629, 'agreement': 0.4675324675324675, 'direct_attack': 0.2765957446808511, 'undercutter_attack': 0.484375, 'partial': 0.08108108108108109}, 'f1': {'support': 0.9007874015748031, 'agreement': 0.6101694915254238, 'direct_attack': 0.25742574257425743, 'undercutter_attack': 0.43356643356643354, 'partial': 0.12765957446808512}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7068965517241379, 'recall': 0.7068965517241379, 'f1': 0.7068965517241379, 'support': None}, 'macro_avg': {'precision': 0.5314696861347061, 'recall': 0.4545094512514726, 'f1': 0.4659217287418006, 'support': None}, 'weighted_avg': {'precision': 0.7020041901737568, 'recall': 0.7068965517241379, 'f1': 0.687907231838573, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3460e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8507462686567164, 'agreement': 0.8536585365853658, 'direct_attack': 0.2631578947368421, 'undercutter_attack': 0.41333333333333333, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9595959595959596, 'agreement': 0.45454545454545453, 'direct_attack': 0.3191489361702128, 'undercutter_attack': 0.484375, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9018987341772151, 'agreement': 0.5932203389830508, 'direct_attack': 0.28846153846153844, 'undercutter_attack': 0.4460431654676259, 'partial': 0.1568627450980392}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7088122605363985, 'recall': 0.7088122605363985, 'f1': 0.7088122605363985, 'support': None}, 'macro_avg': {'precision': 0.5333220638053087, 'recall': 0.46515469168394696, 'f1': 0.47729730443749385, 'support': None}, 'weighted_avg': {'precision': 0.7045910575967651, 'recall': 0.7088122605363985, 'f1': 0.6924334609551257, 'support': None}}
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8511904761904762, 'agreement': 0.8536585365853658, 'direct_attack': 0.2542372881355932, 'undercutter_attack': 0.4246575342465753, 'partial': 0.23076923076923078}, 'recall': {'support': 0.959731543624161, 'agreement': 0.45454545454545453, 'direct_attack': 0.3191489361702128, 'undercutter_attack': 0.484375, 'partial': 0.08333333333333333}, 'f1': {'support': 0.9022082018927444, 'agreement': 0.5932203389830508, 'direct_attack': 0.28301886792452835, 'undercutter_attack': 0.45255474452554745, 'partial': 0.12244897959183675}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 36}, 'micro_avg': {'precision': 0.7088122605363985, 'recall': 0.7088122605363985, 'f1': 0.7088122605363985, 'support': None}, 'macro_avg': {'precision': 0.5229026131854483, 'recall': 0.46022685353463233, 'f1': 0.47069022658354165, 'support': None}, 'weighted_avg': {'precision': 0.7027229813480481, 'recall': 0.7088122605363985, 'f1': 0.6919723447761048, 'support': None}}
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8507462686567164, 'agreement': 0.8536585365853658, 'direct_attack': 0.2542372881355932, 'undercutter_attack': 0.4246575342465753, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9595959595959596, 'agreement': 0.45454545454545453, 'direct_attack': 0.3191489361702128, 'undercutter_attack': 0.484375, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9018987341772151, 'agreement': 0.5932203389830508, 'direct_attack': 0.28301886792452835, 'undercutter_attack': 0.45255474452554745, 'partial': 0.1568627450980392}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7088122605363985, 'recall': 0.7088122605363985, 'f1': 0.7088122605363985, 'support': None}, 'macro_avg': {'precision': 0.5338027826677073, 'recall': 0.46515469168394696, 'f1': 0.47751108614167614, 'support': None}, 'weighted_avg': {'precision': 0.7051762689917629, 'recall': 0.7088122605363985, 'f1': 0.692741766595868, 'support': None}}
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8511904761904762, 'agreement': 0.8536585365853658, 'direct_attack': 0.2711864406779661, 'undercutter_attack': 0.4305555555555556, 'partial': 0.2857142857142857}, 'recall': {'support': 0.959731543624161, 'agreement': 0.45454545454545453, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.49206349206349204, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9022082018927444, 'agreement': 0.5932203389830508, 'direct_attack': 0.30188679245283023, 'undercutter_attack': 0.45925925925925926, 'partial': 0.1568627450980392}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 63, 'partial': 37}, 'micro_avg': {'precision': 0.7126436781609196, 'recall': 0.7126436781609196, 'f1': 0.7126436781609196, 'support': None}, 'macro_avg': {'precision': 0.5384610589447298, 'recall': 0.47097482605122193, 'f1': 0.4826874675371847, 'support': None}, 'weighted_avg': {'precision': 0.7084840239561839, 'recall': 0.7126436781609196, 'f1': 0.6962872498332884, 'support': None}}
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8511904761904762, 'agreement': 0.85, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.4305555555555556, 'partial': 0.2857142857142857}, 'recall': {'support': 0.959731543624161, 'agreement': 0.4473684210526316, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.484375, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9022082018927444, 'agreement': 0.5862068965517242, 'direct_attack': 0.29906542056074764, 'undercutter_attack': 0.4558823529411765, 'partial': 0.1568627450980392}, 'support': {'support': 298, 'agreement': 76, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.710727969348659, 'recall': 0.710727969348659, 'f1': 0.710727969348659, 'support': None}, 'macro_avg': {'precision': 0.5368253968253968, 'recall': 0.46800172093995884, 'f1': 0.4800451234088864, 'support': None}, 'weighted_avg': {'precision': 0.7067338685154777, 'recall': 0.710727969348659, 'f1': 0.6943414467915456, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8511904761904762, 'agreement': 0.8536585365853658, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.43661971830985913, 'partial': 0.2857142857142857}, 'recall': {'support': 0.959731543624161, 'agreement': 0.45454545454545453, 'direct_attack': 0.34782608695652173, 'undercutter_attack': 0.484375, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9022082018927444, 'agreement': 0.5932203389830508, 'direct_attack': 0.30188679245283023, 'undercutter_attack': 0.4592592592592592, 'partial': 0.1568627450980392}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 46, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7126436781609196, 'recall': 0.7126436781609196, 'f1': 0.7126436781609196, 'support': None}, 'macro_avg': {'precision': 0.5387699366933306, 'recall': 0.470917238646849, 'f1': 0.4826874675371847, 'support': None}, 'weighted_avg': {'precision': 0.7091345333941786, 'recall': 0.7126436781609196, 'f1': 0.6965887296547566, 'support': None}}
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2366e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3880e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8507462686567164, 'agreement': 0.8536585365853658, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.4305555555555556, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9595959595959596, 'agreement': 0.45454545454545453, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.484375, 'partial': 0.10810810810810811}, 'f1': {'support': 0.9018987341772151, 'agreement': 0.5932203389830508, 'direct_attack': 0.29906542056074764, 'undercutter_attack': 0.4558823529411765, 'partial': 0.1568627450980392}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.710727969348659, 'recall': 0.710727969348659, 'f1': 0.710727969348659, 'support': None}, 'macro_avg': {'precision': 0.5374682626357179, 'recall': 0.4694100108328831, 'f1': 0.4813859183520458, 'support': None}, 'weighted_avg': {'precision': 0.707018518330336, 'recall': 0.710727969348659, 'f1': 0.6945945537845702, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2399e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0836e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0655e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8507462686567164, 'agreement': 0.8536585365853658, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.43661971830985913, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9595959595959596, 'agreement': 0.45454545454545453, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.484375, 'partial': 0.13513513513513514}, 'f1': {'support': 0.9018987341772151, 'agreement': 0.5932203389830508, 'direct_attack': 0.29906542056074764, 'undercutter_attack': 0.4592592592592592, 'partial': 0.1923076923076923}, 'support': {'support': 297, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 37}, 'micro_avg': {'precision': 0.7126436781609196, 'recall': 0.7126436781609196, 'f1': 0.7126436781609196, 'support': None}, 'macro_avg': {'precision': 0.5482049047103883, 'recall': 0.47481541623828855, 'f1': 0.48915028905759306, 'support': None}, 'weighted_avg': {'precision': 0.7111373136908344, 'recall': 0.7126436781609196, 'f1': 0.6975209619284677, 'support': None}}
Loss: tensor(9.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7857e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6589e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9363e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6040e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9138e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0844e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4667e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3979e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8537313432835821, 'agreement': 0.8536585365853658, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.43661971830985913, 'partial': 0.3333333333333333}, 'recall': {'support': 0.959731543624161, 'agreement': 0.45454545454545453, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.484375, 'partial': 0.1388888888888889}, 'f1': {'support': 0.9036334913112164, 'agreement': 0.5932203389830508, 'direct_attack': 0.29906542056074764, 'undercutter_attack': 0.4592592592592592, 'partial': 0.19607843137254902}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 36}, 'micro_avg': {'precision': 0.7145593869731801, 'recall': 0.7145593869731801, 'f1': 0.7145593869731802, 'support': None}, 'macro_avg': {'precision': 0.5488019196357614, 'recall': 0.4755932837946796, 'f1': 0.4902513882973647, 'support': None}, 'weighted_avg': {'precision': 0.7138326492734579, 'recall': 0.7145593869731801, 'f1': 0.7001307229900323, 'support': None}}
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5410e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9968e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5402e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1490e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3561e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7808e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6032e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2391e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3307e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1785e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0885e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8152e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8160e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3586e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1793e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0997e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8127e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8557e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3610e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8511904761904762, 'agreement': 0.8536585365853658, 'direct_attack': 0.26666666666666666, 'undercutter_attack': 0.44285714285714284, 'partial': 0.3333333333333333}, 'recall': {'support': 0.959731543624161, 'agreement': 0.45454545454545453, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.484375, 'partial': 0.1388888888888889}, 'f1': {'support': 0.9022082018927444, 'agreement': 0.5932203389830508, 'direct_attack': 0.29906542056074764, 'undercutter_attack': 0.4626865671641791, 'partial': 0.19607843137254902}, 'support': {'support': 298, 'agreement': 77, 'direct_attack': 47, 'undercutter_attack': 64, 'partial': 36}, 'micro_avg': {'precision': 0.7145593869731801, 'recall': 0.7145593869731801, 'f1': 0.7145593869731802, 'support': None}, 'macro_avg': {'precision': 0.549541231126597, 'recall': 0.4755932837946796, 'f1': 0.4906517919946543, 'support': None}, 'weighted_avg': {'precision': 0.7131468576590528, 'recall': 0.7145593869731801, 'f1': 0.6997372583525041, 'support': None}}
	Train size: 50 Test size: 50


		-------------RUN 1-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.6386292834890965, 'agreement': 0.0, 'direct_attack': 0.07742878013148284, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.1967370441458733, 'agreement': 0.0, 'direct_attack': 0.8346456692913385, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.3008070432868672, 'agreement': 0.0, 'direct_attack': 0.14171122994652408, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.18402366863905326, 'recall': 0.18402366863905326, 'f1': 0.18402366863905326, 'support': None}, 'macro_avg': {'precision': 0.14321161272411587, 'recall': 0.20627654268744236, 'f1': 0.08850365464667825, 'support': None}, 'weighted_avg': {'precision': 0.39957702276469637, 'recall': 0.18402366863905326, 'f1': 0.19611731675036934, 'support': None}}
Loss: tensor(0.7465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8450, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.6463683052090976, 'agreement': 0.43902439024390244, 'direct_attack': 0.10344827586206896, 'undercutter_attack': 0.1875, 'partial': 0.022988505747126436}, 'recall': {'support': 0.845489443378119, 'agreement': 0.0821917808219178, 'direct_attack': 0.07086614173228346, 'undercutter_attack': 0.11538461538461539, 'partial': 0.016666666666666666}, 'f1': {'support': 0.7326403326403327, 'agreement': 0.13846153846153844, 'direct_attack': 0.08411214953271028, 'undercutter_attack': 0.14285714285714285, 'partial': 0.019323671497584544}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.5508875739644971, 'recall': 0.5508875739644971, 'f1': 0.5508875739644971, 'support': None}, 'macro_avg': {'precision': 0.27986589541243906, 'recall': 0.22611972959672047, 'f1': 0.22347896699786177, 'support': None}, 'weighted_avg': {'precision': 0.4850199214292499, 'recall': 0.5508875739644971, 'f1': 0.49274283260631235, 'support': None}}
Loss: tensor(0.5633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2820, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.863517060367454, 'agreement': 0.3341645885286783, 'direct_attack': 0.125, 'undercutter_attack': 0.18793503480278423, 'partial': 0.053763440860215055}, 'recall': {'support': 0.31573896353166986, 'agreement': 0.6118721461187214, 'direct_attack': 0.3779527559055118, 'undercutter_attack': 0.44505494505494503, 'partial': 0.041666666666666664}, 'f1': {'support': 0.4624033731553057, 'agreement': 0.432258064516129, 'direct_attack': 0.18786692759295498, 'undercutter_attack': 0.2642740619902121, 'partial': 0.046948356807511735}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.35325443786982247, 'recall': 0.35325443786982247, 'f1': 0.35325443786982247, 'support': None}, 'macro_avg': {'precision': 0.3128760249118263, 'recall': 0.35845709545550297, 'f1': 0.2787501568124227, 'support': None}, 'weighted_avg': {'precision': 0.6091701840402368, 'recall': 0.35325443786982247, 'f1': 0.3870293567220628, 'support': None}}
Loss: tensor(0.9231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6825, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.764760147601476, 'agreement': 0.5125628140703518, 'direct_attack': 0.1875, 'undercutter_attack': 0.3275862068965517, 'partial': 0.1320754716981132}, 'recall': {'support': 0.7955854126679462, 'agreement': 0.4657534246575342, 'direct_attack': 0.023622047244094488, 'undercutter_attack': 0.4175824175824176, 'partial': 0.175}, 'f1': {'support': 0.779868297271872, 'agreement': 0.48803827751196166, 'direct_attack': 0.04195804195804195, 'undercutter_attack': 0.3671497584541063, 'partial': 0.15053763440860213}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6100591715976331, 'recall': 0.6100591715976331, 'f1': 0.6100591715976331, 'support': None}, 'macro_avg': {'precision': 0.3848969280532985, 'recall': 0.3755086604303985, 'f1': 0.3655104019209168, 'support': None}, 'weighted_avg': {'precision': 0.596694423870468, 'recall': 0.6100591715976331, 'f1': 0.5974660307862492, 'support': None}}
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7165653495440729, 'agreement': 0.43373493975903615, 'direct_attack': 0.25, 'undercutter_attack': 0.5454545454545454, 'partial': 0.0}, 'recall': {'support': 0.9049904030710173, 'agreement': 0.4931506849315068, 'direct_attack': 0.18110236220472442, 'undercutter_attack': 0.06593406593406594, 'partial': 0.0}, 'f1': {'support': 0.799830364715861, 'agreement': 0.4615384615384615, 'direct_attack': 0.21004566210045664, 'undercutter_attack': 0.11764705882352941, 'partial': 0.0}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.642603550295858, 'recall': 0.642603550295858, 'f1': 0.642603550295858, 'support': None}, 'macro_avg': {'precision': 0.3891509669515309, 'recall': 0.3290355032282629, 'f1': 0.31781230943566174, 'support': None}, 'weighted_avg': {'precision': 0.5755454279910533, 'recall': 0.642603550295858, 'f1': 0.5814128561559117, 'support': None}}
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7272036474164134, 'agreement': 0.5783132530120482, 'direct_attack': 0.22429906542056074, 'undercutter_attack': 0.3160919540229885, 'partial': 0.2}, 'recall': {'support': 0.918426103646833, 'agreement': 0.2191780821917808, 'direct_attack': 0.1889763779527559, 'undercutter_attack': 0.3021978021978022, 'partial': 0.016666666666666666}, 'f1': {'support': 0.8117048346055981, 'agreement': 0.3178807947019867, 'direct_attack': 0.2051282051282051, 'undercutter_attack': 0.30898876404494374, 'partial': 0.03076923076923077}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.642603550295858, 'recall': 0.642603550295858, 'f1': 0.642603550295858, 'support': None}, 'macro_avg': {'precision': 0.40918158397440213, 'recall': 0.32908900653116774, 'f1': 0.3348943658499929, 'support': None}, 'weighted_avg': {'precision': 0.5884091834071813, 'recall': 0.642603550295858, 'f1': 0.5925395718926258, 'support': None}}
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.797244094488189, 'agreement': 0.4585152838427948, 'direct_attack': 0.2782608695652174, 'undercutter_attack': 0.2961783439490446, 'partial': 0.125}, 'recall': {'support': 0.7773512476007678, 'agreement': 0.4794520547945205, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.510989010989011, 'partial': 0.016666666666666666}, 'f1': {'support': 0.7871720116618076, 'agreement': 0.46875, 'direct_attack': 0.26446280991735543, 'undercutter_attack': 0.37499999999999994, 'partial': 0.029411764705882353}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6165680473372781, 'recall': 0.6165680473372781, 'f1': 0.6165680473372781, 'support': None}, 'macro_avg': {'precision': 0.39103971836904916, 'recall': 0.4072854967975948, 'f1': 0.38495931725700905, 'support': None}, 'weighted_avg': {'precision': 0.6126549009773811, 'recall': 0.6165680473372781, 'f1': 0.6084353105182329, 'support': None}}
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7408585055643879, 'agreement': 0.4887640449438202, 'direct_attack': 0.3063063063063063, 'undercutter_attack': 0.41975308641975306, 'partial': 0.25806451612903225}, 'recall': {'support': 0.8944337811900192, 'agreement': 0.3972602739726027, 'direct_attack': 0.2677165354330709, 'undercutter_attack': 0.18681318681318682, 'partial': 0.13333333333333333}, 'f1': {'support': 0.8104347826086956, 'agreement': 0.4382871536523929, 'direct_attack': 0.28571428571428575, 'undercutter_attack': 0.2585551330798479, 'partial': 0.17582417582417584}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6526627218934912, 'recall': 0.6526627218934912, 'f1': 0.6526627218934912, 'support': None}, 'macro_avg': {'precision': 0.44274929187266, 'recall': 0.3759114221484426, 'f1': 0.3937631061758796, 'support': None}, 'weighted_avg': {'precision': 0.606673132074301, 'recall': 0.6526627218934912, 'f1': 0.618283774990108, 'support': None}}
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7491830065359477, 'agreement': 0.4945054945054945, 'direct_attack': 0.3137254901960784, 'undercutter_attack': 0.3767123287671233, 'partial': 0.25}, 'recall': {'support': 0.8800383877159309, 'agreement': 0.410958904109589, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3021978021978022, 'partial': 0.075}, 'f1': {'support': 0.8093556928508385, 'agreement': 0.44887780548628425, 'direct_attack': 0.2794759825327511, 'undercutter_attack': 0.33536585365853655, 'partial': 0.11538461538461538}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6526627218934912, 'recall': 0.6526627218934912, 'f1': 0.6526627218934912, 'support': None}, 'macro_avg': {'precision': 0.4368252640009288, 'recall': 0.384032719592066, 'f1': 0.39769198998260513, 'support': None}, 'weighted_avg': {'precision': 0.6078995131347216, 'recall': 0.6526627218934912, 'f1': 0.6225024025714418, 'support': None}}
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7493877551020408, 'agreement': 0.4918918918918919, 'direct_attack': 0.3191489361702128, 'undercutter_attack': 0.3691275167785235, 'partial': 0.24324324324324326}, 'recall': {'support': 0.8809980806142035, 'agreement': 0.4155251141552511, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3021978021978022, 'partial': 0.075}, 'f1': {'support': 0.8098808998676665, 'agreement': 0.4504950495049505, 'direct_attack': 0.27149321266968324, 'undercutter_attack': 0.3323262839879154, 'partial': 0.11464968152866242}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6526627218934912, 'recall': 0.6526627218934912, 'f1': 0.6526627218934912, 'support': None}, 'macro_avg': {'precision': 0.43455986863718243, 'recall': 0.38198829388164035, 'f1': 0.39576902551177556, 'support': None}, 'weighted_avg': {'precision': 0.6067980338918038, 'recall': 0.6526627218934912, 'f1': 0.6220563887467351, 'support': None}}
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7493877551020408, 'agreement': 0.4919786096256685, 'direct_attack': 0.3157894736842105, 'undercutter_attack': 0.38513513513513514, 'partial': 0.2571428571428571}, 'recall': {'support': 0.8809980806142035, 'agreement': 0.4200913242009132, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8098808998676665, 'agreement': 0.4532019704433498, 'direct_attack': 0.27027027027027023, 'undercutter_attack': 0.3454545454545455, 'partial': 0.11612903225806451}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6544378698224852, 'recall': 0.6544378698224852, 'f1': 0.6544378698224852, 'support': None}, 'macro_avg': {'precision': 0.4398867661379824, 'recall': 0.38509933808857494, 'f1': 0.3989873436587793, 'support': None}, 'weighted_avg': {'precision': 0.6092676668248402, 'recall': 0.6544378698224852, 'f1': 0.6238341210989476, 'support': None}}
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7497962510187449, 'agreement': 0.4946236559139785, 'direct_attack': 0.3225806451612903, 'undercutter_attack': 0.3825503355704698, 'partial': 0.2571428571428571}, 'recall': {'support': 0.8829174664107485, 'agreement': 0.4200913242009132, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8109299250771265, 'agreement': 0.454320987654321, 'direct_attack': 0.2727272727272727, 'undercutter_attack': 0.3444108761329305, 'partial': 0.11612903225806451}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6556213017751479, 'recall': 0.6556213017751479, 'f1': 0.6556213017751479, 'support': None}, 'macro_avg': {'precision': 0.44133874896146813, 'recall': 0.3854832152478839, 'f1': 0.3997036187699431, 'support': None}, 'weighted_avg': {'precision': 0.6100942722326306, 'recall': 0.6556213017751479, 'f1': 0.6246981687515899, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7493897477624084, 'agreement': 0.4946236559139785, 'direct_attack': 0.32608695652173914, 'undercutter_attack': 0.3877551020408163, 'partial': 0.25}, 'recall': {'support': 0.8838771593090211, 'agreement': 0.4200913242009132, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8110964332892998, 'agreement': 0.454320987654321, 'direct_attack': 0.273972602739726, 'undercutter_attack': 0.3465045592705167, 'partial': 0.11538461538461538}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6562130177514793, 'recall': 0.6562130177514793, 'f1': 0.6562130177514793, 'support': None}, 'macro_avg': {'precision': 0.4415710924477885, 'recall': 0.3856751538275385, 'f1': 0.40025583966769573, 'support': None}, 'weighted_avg': {'precision': 0.6101604555404025, 'recall': 0.6562130177514793, 'f1': 0.6250670319319999, 'support': None}}
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7493897477624084, 'agreement': 0.4946236559139785, 'direct_attack': 0.32608695652173914, 'undercutter_attack': 0.3877551020408163, 'partial': 0.25}, 'recall': {'support': 0.8838771593090211, 'agreement': 0.4200913242009132, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8110964332892998, 'agreement': 0.454320987654321, 'direct_attack': 0.273972602739726, 'undercutter_attack': 0.3465045592705167, 'partial': 0.11538461538461538}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6562130177514793, 'recall': 0.6562130177514793, 'f1': 0.6562130177514793, 'support': None}, 'macro_avg': {'precision': 0.4415710924477885, 'recall': 0.3856751538275385, 'f1': 0.40025583966769573, 'support': None}, 'weighted_avg': {'precision': 0.6101604555404025, 'recall': 0.6562130177514793, 'f1': 0.6250670319319999, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7502034174125305, 'agreement': 0.4946236559139785, 'direct_attack': 0.32967032967032966, 'undercutter_attack': 0.38513513513513514, 'partial': 0.25}, 'recall': {'support': 0.8848368522072937, 'agreement': 0.4200913242009132, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8119771025979744, 'agreement': 0.454320987654321, 'direct_attack': 0.27522935779816515, 'undercutter_attack': 0.3454545454545455, 'partial': 0.11538461538461538}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6568047337278107, 'recall': 0.6568047337278107, 'f1': 0.6568047337278107, 'support': None}, 'macro_avg': {'precision': 0.44192650762639474, 'recall': 0.385867092407193, 'f1': 0.4004733217779243, 'support': None}, 'weighted_avg': {'precision': 0.6106492710365351, 'recall': 0.6568047337278107, 'f1': 0.6255913886169431, 'support': None}}
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7493917274939172, 'agreement': 0.4972677595628415, 'direct_attack': 0.34065934065934067, 'undercutter_attack': 0.38513513513513514, 'partial': 0.2571428571428571}, 'recall': {'support': 0.8867562380038387, 'agreement': 0.4155251141552511, 'direct_attack': 0.2440944881889764, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8123076923076923, 'agreement': 0.4527363184079602, 'direct_attack': 0.28440366972477066, 'undercutter_attack': 0.3454545454545455, 'partial': 0.11612903225806451}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6579881656804734, 'recall': 0.6579881656804734, 'f1': 0.6579881656804734, 'support': None}, 'macro_avg': {'precision': 0.4459193639988183, 'recall': 0.3869125307069759, 'f1': 0.40220625163060664, 'support': None}, 'weighted_avg': {'precision': 0.6118244337919514, 'recall': 0.6579881656804734, 'f1': 0.6263321575826625, 'support': None}}
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7502027575020276, 'agreement': 0.4972677595628415, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.3904109589041096, 'partial': 0.2571428571428571}, 'recall': {'support': 0.8877159309021113, 'agreement': 0.4155251141552511, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8131868131868131, 'agreement': 0.4527363184079602, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.3475609756097561, 'partial': 0.11612903225806451}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.659171597633136, 'recall': 0.659171597633136, 'f1': 0.659171597633136, 'support': None}, 'macro_avg': {'precision': 0.44782207092344245, 'recall': 0.3886792724362367, 'f1': 0.404104446074337, 'support': None}, 'weighted_avg': {'precision': 0.6131501625859459, 'recall': 0.659171597633136, 'f1': 0.6275899106801187, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7493936944219887, 'agreement': 0.4972067039106145, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.3931034482758621, 'partial': 0.2571428571428571}, 'recall': {'support': 0.8896353166986565, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3131868131868132, 'partial': 0.075}, 'f1': {'support': 0.8135146994295744, 'agreement': 0.4472361809045226, 'direct_attack': 0.28959276018099545, 'undercutter_attack': 0.34862385321100914, 'partial': 0.11612903225806451}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.659171597633136, 'recall': 0.659171597633136, 'f1': 0.659171597633136, 'support': None}, 'macro_avg': {'precision': 0.4474544471332432, 'recall': 0.38723666557728087, 'f1': 0.40301930519683316, 'support': None}, 'weighted_avg': {'precision': 0.6126582903790995, 'recall': 0.659171597633136, 'f1': 0.6270948793621686, 'support': None}}
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7495961227786753, 'agreement': 0.49444444444444446, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.3916083916083916, 'partial': 0.2777777777777778}, 'recall': {'support': 0.8905950095969289, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3076923076923077, 'partial': 0.08333333333333333}, 'f1': {'support': 0.8140350877192981, 'agreement': 0.4461152882205514, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.3446153846153846, 'partial': 0.1282051282051282}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6597633136094675, 'recall': 0.6597633136094675, 'f1': 0.6597633136094675, 'support': None}, 'macro_avg': {'precision': 0.45150255162293307, 'recall': 0.387996369724701, 'f1': 0.40477599593389063, 'support': None}, 'weighted_avg': {'precision': 0.614004425210625, 'recall': 0.6597633136094675, 'f1': 0.627795194943124, 'support': None}}
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7489911218724778, 'agreement': 0.4972067039106145, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.39436619718309857, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8905950095969289, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3076923076923077, 'partial': 0.08333333333333333}, 'f1': {'support': 0.8136782113108285, 'agreement': 0.4472361809045226, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.34567901234567905, 'partial': 0.12738853503184713}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6597633136094675, 'recall': 0.6597633136094675, 'f1': 0.6597633136094675, 'support': None}, 'macro_avg': {'precision': 0.45098406294836746, 'recall': 0.387996369724701, 'f1': 0.40497820610039365, 'support': None}, 'weighted_avg': {'precision': 0.6137532675730685, 'recall': 0.6597633136094675, 'f1': 0.6277769697042389, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.748792270531401, 'agreement': 0.5028248587570622, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.3971631205673759, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8925143953934741, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3076923076923077, 'partial': 0.08333333333333333}, 'f1': {'support': 0.8143607705779334, 'agreement': 0.4494949494949495, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.34674922600619196, 'partial': 0.12738853503184713}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6609467455621302, 'recall': 0.6609467455621302, 'f1': 0.6609467455621302, 'support': None}, 'macro_avg': {'precision': 0.45262730832629716, 'recall': 0.38838024688400996, 'f1': 0.40578051440400265, 'support': None}, 'weighted_avg': {'precision': 0.6146599024073338, 'recall': 0.6609467455621302, 'f1': 0.6286057720497064, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7489943684633951, 'agreement': 0.5056818181818182, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.3971631205673759, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8934740882917467, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3076923076923077, 'partial': 0.08333333333333333}, 'f1': {'support': 0.8148796498905908, 'agreement': 0.450632911392405, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.34674922600619196, 'partial': 0.12738853503184713}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6615384615384615, 'recall': 0.6615384615384615, 'f1': 0.6615384615384615, 'support': None}, 'macro_avg': {'precision': 0.4532391197976472, 'recall': 0.38857218546366445, 'f1': 0.40611188264602516, 'support': None}, 'weighted_avg': {'precision': 0.6151547309038777, 'recall': 0.6615384615384615, 'f1': 0.6290731601558199, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7489943684633951, 'agreement': 0.5056818181818182, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.3971631205673759, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8934740882917467, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3076923076923077, 'partial': 0.08333333333333333}, 'f1': {'support': 0.8148796498905908, 'agreement': 0.450632911392405, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.34674922600619196, 'partial': 0.12738853503184713}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6615384615384615, 'recall': 0.6615384615384615, 'f1': 0.6615384615384615, 'support': None}, 'macro_avg': {'precision': 0.4532391197976472, 'recall': 0.38857218546366445, 'f1': 0.40611188264602516, 'support': None}, 'weighted_avg': {'precision': 0.6151547309038777, 'recall': 0.6615384615384615, 'f1': 0.6290731601558199, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7483922829581994, 'agreement': 0.5056818181818182, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.39285714285714285, 'partial': 0.2702702702702703}, 'recall': {'support': 0.8934740882917467, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3021978021978022, 'partial': 0.08333333333333333}, 'f1': {'support': 0.8145231846019247, 'agreement': 0.450632911392405, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.3416149068322981, 'partial': 0.12738853503184713}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6609467455621302, 'recall': 0.6609467455621302, 'f1': 0.6609467455621302, 'support': None}, 'macro_avg': {'precision': 0.45225750715456137, 'recall': 0.38747328436476336, 'f1': 0.40501372575351324, 'support': None}, 'weighted_avg': {'precision': 0.6143197835431227, 'recall': 0.6609467455621302, 'f1': 0.628300448368578, 'support': None}}
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7477911646586345, 'agreement': 0.5028571428571429, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.39568345323741005, 'partial': 0.2894736842105263}, 'recall': {'support': 0.8934740882917467, 'agreement': 0.4018264840182648, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3021978021978022, 'partial': 0.09166666666666666}, 'f1': {'support': 0.8141670310450371, 'agreement': 0.4467005076142132, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.3426791277258566, 'partial': 0.13924050632911392}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6609467455621302, 'recall': 0.6609467455621302, 'f1': 0.6609467455621302, 'support': None}, 'macro_avg': {'precision': 0.455978293293818, 'recall': 0.3882267090222976, 'f1': 0.4067392527246623, 'support': None}, 'weighted_avg': {'precision': 0.6152510433051278, 'recall': 0.6609467455621302, 'f1': 0.6285274402766247, 'support': None}}
Loss: tensor(8.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4691e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7477911646586345, 'agreement': 0.5, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.39568345323741005, 'partial': 0.2972972972972973}, 'recall': {'support': 0.8934740882917467, 'agreement': 0.4018264840182648, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3021978021978022, 'partial': 0.09166666666666666}, 'f1': {'support': 0.8141670310450371, 'agreement': 0.44556962025316454, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.3426791277258566, 'partial': 0.14012738853503184}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6609467455621302, 'recall': 0.6609467455621302, 'f1': 0.6609467455621302, 'support': None}, 'macro_avg': {'precision': 0.45697158733974363, 'recall': 0.3882267090222976, 'f1': 0.4066904516936362, 'support': None}, 'weighted_avg': {'precision': 0.6154363209883812, 'recall': 0.6609467455621302, 'f1': 0.6284438672190154, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7477911646586345, 'agreement': 0.5, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.39855072463768115, 'partial': 0.3157894736842105}, 'recall': {'support': 0.8934740882917467, 'agreement': 0.4018264840182648, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3021978021978022, 'partial': 0.1}, 'f1': {'support': 0.8141670310450371, 'agreement': 0.44556962025316454, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.34374999999999994, 'partial': 0.15189873417721522}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6615384615384615, 'recall': 0.6615384615384615, 'f1': 0.6615384615384615, 'support': None}, 'macro_avg': {'precision': 0.4612434768971806, 'recall': 0.3898933756889643, 'f1': 0.40925889527690157, 'support': None}, 'weighted_avg': {'precision': 0.6170581580068895, 'recall': 0.6615384615384615, 'f1': 0.6293950271189894, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7475961538461539, 'agreement': 0.5, 'direct_attack': 0.35555555555555557, 'undercutter_attack': 0.39705882352941174, 'partial': 0.3157894736842105}, 'recall': {'support': 0.8953934740882917, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.2967032967032967, 'partial': 0.1}, 'f1': {'support': 0.8148471615720524, 'agreement': 0.44836272040302266, 'direct_attack': 0.29493087557603687, 'undercutter_attack': 0.33962264150943394, 'partial': 0.15189873417721522}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6627218934911243, 'recall': 0.6627218934911243, 'f1': 0.6627218934911243, 'support': None}, 'macro_avg': {'precision': 0.4632000013230663, 'recall': 0.39009159375850466, 'f1': 0.4099324266475522, 'support': None}, 'weighted_avg': {'precision': 0.6176391660282284, 'recall': 0.6627218934911243, 'f1': 0.6300340640121184, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7479935794542536, 'agreement': 0.5028248587570622, 'direct_attack': 0.3516483516483517, 'undercutter_attack': 0.4057971014492754, 'partial': 0.3157894736842105}, 'recall': {'support': 0.8944337811900192, 'agreement': 0.4063926940639269, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3076923076923077, 'partial': 0.1}, 'f1': {'support': 0.8146853146853147, 'agreement': 0.4494949494949495, 'direct_attack': 0.2935779816513761, 'undercutter_attack': 0.35000000000000003, 'partial': 0.15189873417721522}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6633136094674557, 'recall': 0.6633136094674557, 'f1': 0.6633136094674557, 'support': None}, 'macro_avg': {'precision': 0.4648106729986307, 'recall': 0.39209745737665236, 'f1': 0.41193139600177114, 'support': None}, 'weighted_avg': {'precision': 0.618897694570617, 'recall': 0.6633136094674557, 'f1': 0.6310968897115281, 'support': None}}
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7479935794542536, 'agreement': 0.5, 'direct_attack': 0.34408602150537637, 'undercutter_attack': 0.40145985401459855, 'partial': 0.3157894736842105}, 'recall': {'support': 0.8944337811900192, 'agreement': 0.4018264840182648, 'direct_attack': 0.25196850393700787, 'undercutter_attack': 0.3021978021978022, 'partial': 0.1}, 'f1': {'support': 0.8146853146853147, 'agreement': 0.44556962025316454, 'direct_attack': 0.29090909090909095, 'undercutter_attack': 0.3448275862068965, 'partial': 0.15189873417721522}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6621301775147929, 'recall': 0.6621301775147929, 'f1': 0.6621301775147929, 'support': None}, 'macro_avg': {'precision': 0.46186578573168785, 'recall': 0.3900853142686188, 'f1': 0.40957806924633644, 'support': None}, 'weighted_avg': {'precision': 0.6174962513581522, 'recall': 0.6621301775147929, 'f1': 0.6298306319964003, 'support': None}}
Loss: tensor(8.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7475961538461539, 'agreement': 0.5028248587570622, 'direct_attack': 0.32967032967032966, 'undercutter_attack': 0.40441176470588236, 'partial': 0.3157894736842105}, 'recall': {'support': 0.8953934740882917, 'agreement': 0.4063926940639269, 'direct_attack': 0.23622047244094488, 'undercutter_attack': 0.3021978021978022, 'partial': 0.1}, 'f1': {'support': 0.8148471615720524, 'agreement': 0.4494949494949495, 'direct_attack': 0.27522935779816515, 'undercutter_attack': 0.3459119496855346, 'partial': 0.15189873417721522}, 'support': {'support': 1042, 'agreement': 219, 'direct_attack': 127, 'undercutter_attack': 182, 'partial': 120}, 'micro_avg': {'precision': 0.6621301775147929, 'recall': 0.6621301775147929, 'f1': 0.6621301775147929, 'support': None}, 'macro_avg': {'precision': 0.46005851613272775, 'recall': 0.3880408885581932, 'f1': 0.4074764305455834, 'support': None}, 'weighted_avg': {'precision': 0.6168518616936075, 'recall': 0.6621301775147929, 'f1': 0.6293775666756644, 'support': None}}


		-------------RUN 2-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.39655172413793105, 'agreement': 0.09347442680776014, 'direct_attack': 0.07142857142857142, 'undercutter_attack': 0.1056701030927835, 'partial': 0.05605786618444846}, 'recall': {'support': 0.022885572139303482, 'agreement': 0.28191489361702127, 'direct_attack': 0.017699115044247787, 'undercutter_attack': 0.21693121693121692, 'partial': 0.31313131313131315}, 'f1': {'support': 0.04327375352775164, 'agreement': 0.14039735099337747, 'direct_attack': 0.028368794326241134, 'undercutter_attack': 0.14211438474870017, 'partial': 0.0950920245398773}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.09410288582183186, 'recall': 0.09410288582183186, 'f1': 0.09410288582183186, 'support': None}, 'macro_avg': {'precision': 0.1446365383302989, 'recall': 0.1705124221726205, 'f1': 0.08984926162718954, 'support': None}, 'weighted_avg': {'precision': 0.2821207539565274, 'recall': 0.09410288582183186, 'f1': 0.06860992922707829, 'support': None}}
Loss: tensor(2.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6789, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7174095878889823, 'agreement': 0.7272727272727273, 'direct_attack': 0.11363636363636363, 'undercutter_attack': 0.21238938053097345, 'partial': 0.0}, 'recall': {'support': 0.8487562189054726, 'agreement': 0.0851063829787234, 'direct_attack': 0.04424778761061947, 'undercutter_attack': 0.38095238095238093, 'partial': 0.0}, 'f1': {'support': 0.7775752051048314, 'agreement': 0.15238095238095237, 'direct_attack': 0.06369426751592357, 'undercutter_attack': 0.2727272727272727, 'partial': 0.0}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.5934755332496863, 'recall': 0.5934755332496863, 'f1': 0.5934755332496863, 'support': None}, 'macro_avg': {'precision': 0.3541416118658093, 'recall': 0.2718125540894393, 'f1': 0.253275539545796, 'support': None}, 'weighted_avg': {'precision': 0.5713340091386219, 'recall': 0.5934755332496863, 'f1': 0.5450775451397293, 'support': None}}
Loss: tensor(2.8545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5591, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9315726290516206, 'agreement': 0.6206896551724138, 'direct_attack': 0.15450643776824036, 'undercutter_attack': 0.75, 'partial': 0.0}, 'recall': {'support': 0.7713717693836978, 'agreement': 0.1925133689839572, 'direct_attack': 0.9557522123893806, 'undercutter_attack': 0.015873015873015872, 'partial': 0.0}, 'f1': {'support': 0.8439369222403479, 'agreement': 0.29387755102040813, 'direct_attack': 0.2660098522167488, 'undercutter_attack': 0.031088082901554404, 'partial': 0.0}, 'support': {'support': 1006, 'agreement': 187, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.5790464240903388, 'recall': 0.5790464240903388, 'f1': 0.5790464240903388, 'support': None}, 'macro_avg': {'precision': 0.4913537443984549, 'recall': 0.3871020733260103, 'f1': 0.28698248167581186, 'support': None}, 'weighted_avg': {'precision': 0.7606275143105287, 'recall': 0.5790464240903388, 'f1': 0.5896426642305475, 'support': None}}
Loss: tensor(0.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6116, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8897266729500471, 'agreement': 0.4954954954954955, 'direct_attack': 0.25, 'undercutter_attack': 0.49825783972125437, 'partial': 0.3103448275862069}, 'recall': {'support': 0.9393034825870646, 'agreement': 0.2925531914893617, 'direct_attack': 0.10619469026548672, 'undercutter_attack': 0.7566137566137566, 'partial': 0.2727272727272727}, 'f1': {'support': 0.9138431752178122, 'agreement': 0.36789297658862874, 'direct_attack': 0.14906832298136646, 'undercutter_attack': 0.6008403361344538, 'partial': 0.2903225806451613}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7409033877038896, 'recall': 0.7409033877038896, 'f1': 0.7409033877038896, 'support': None}, 'macro_avg': {'precision': 0.4887649671506008, 'recall': 0.4734784787365884, 'f1': 0.46439347831348454, 'support': None}, 'weighted_avg': {'precision': 0.7154788764782322, 'recall': 0.7409033877038896, 'f1': 0.7193988395249313, 'support': None}}
Loss: tensor(5.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9079878665318504, 'agreement': 0.37358490566037733, 'direct_attack': 0.17, 'undercutter_attack': 0.45794392523364486, 'partial': 0.4230769230769231}, 'recall': {'support': 0.8926441351888668, 'agreement': 0.526595744680851, 'direct_attack': 0.1504424778761062, 'undercutter_attack': 0.5212765957446809, 'partial': 0.1111111111111111}, 'f1': {'support': 0.900250626566416, 'agreement': 0.43708609271523174, 'direct_attack': 0.1596244131455399, 'undercutter_attack': 0.4875621890547263, 'partial': 0.176}, 'support': {'support': 1006, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 188, 'partial': 99}, 'micro_avg': {'precision': 0.704516938519448, 'recall': 0.704516938519448, 'f1': 0.704516938519448, 'support': None}, 'macro_avg': {'precision': 0.4665187241005591, 'recall': 0.44041401292032323, 'f1': 0.4321046642963828, 'support': None}, 'weighted_avg': {'precision': 0.7094465679571725, 'recall': 0.704516938519448, 'f1': 0.6994652233274861, 'support': None}}
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9190573770491803, 'agreement': 0.44933920704845814, 'direct_attack': 0.26506024096385544, 'undercutter_attack': 0.5, 'partial': 0.4444444444444444}, 'recall': {'support': 0.8925373134328358, 'agreement': 0.5425531914893617, 'direct_attack': 0.19469026548672566, 'undercutter_attack': 0.7195767195767195, 'partial': 0.16161616161616163}, 'f1': {'support': 0.9056032306915699, 'agreement': 0.4915662650602409, 'direct_attack': 0.22448979591836732, 'undercutter_attack': 0.5900216919739696, 'partial': 0.23703703703703705}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7358845671267252, 'recall': 0.7358845671267252, 'f1': 0.7358845671267252, 'support': None}, 'macro_avg': {'precision': 0.5155802539011877, 'recall': 0.5021947303203609, 'f1': 0.48974360413623697, 'support': None}, 'weighted_avg': {'precision': 0.7381306412098193, 'recall': 0.7358845671267252, 'f1': 0.729544427895154, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9153924566768603, 'agreement': 0.5197368421052632, 'direct_attack': 0.1919191919191919, 'undercutter_attack': 0.48360655737704916, 'partial': 0.3157894736842105}, 'recall': {'support': 0.8935323383084577, 'agreement': 0.42021276595744683, 'direct_attack': 0.336283185840708, 'undercutter_attack': 0.6243386243386243, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9043303121852971, 'agreement': 0.4647058823529412, 'direct_attack': 0.24437299035369778, 'undercutter_attack': 0.5450346420323325, 'partial': 0.1016949152542373}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7145545796737767, 'recall': 0.7145545796737767, 'f1': 0.7145545796737767, 'support': None}, 'macro_avg': {'precision': 0.485288904352515, 'recall': 0.46699459501025953, 'f1': 0.45202774843570126, 'support': None}, 'weighted_avg': {'precision': 0.729003520201946, 'recall': 0.7145545796737767, 'f1': 0.7132435141109316, 'support': None}}
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.884272997032641, 'agreement': 0.45077720207253885, 'direct_attack': 0.2604166666666667, 'undercutter_attack': 0.47017543859649125, 'partial': 0.6666666666666666}, 'recall': {'support': 0.8886679920477137, 'agreement': 0.4627659574468085, 'direct_attack': 0.22123893805309736, 'undercutter_attack': 0.708994708994709, 'partial': 0.061224489795918366}, 'f1': {'support': 0.886465047099653, 'agreement': 0.4566929133858268, 'direct_attack': 0.23923444976076555, 'undercutter_attack': 0.5654008438818565, 'partial': 0.11214953271028037}, 'support': {'support': 1006, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 98}, 'micro_avg': {'precision': 0.7189460476787954, 'recall': 0.7189460476787954, 'f1': 0.7189460476787954, 'support': None}, 'macro_avg': {'precision': 0.5464617942070008, 'recall': 0.46857841726764937, 'f1': 0.4519885573676764, 'support': None}, 'weighted_avg': {'precision': 0.726441859200676, 'recall': 0.7189460476787954, 'f1': 0.7042202080433069, 'support': None}}
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9237089201877934, 'agreement': 0.5096774193548387, 'direct_attack': 0.1554054054054054, 'undercutter_attack': 0.4117647058823529, 'partial': 1.0}, 'recall': {'support': 0.7830845771144279, 'agreement': 0.42021276595744683, 'direct_attack': 0.40707964601769914, 'undercutter_attack': 0.6296296296296297, 'partial': 0.020202020202020204}, 'f1': {'support': 0.84760366182014, 'agreement': 0.4606413994169096, 'direct_attack': 0.22493887530562345, 'undercutter_attack': 0.49790794979079495, 'partial': 0.03960396039603961}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.6480552070263488, 'recall': 0.6480552070263488, 'f1': 0.6480552070263488, 'support': None}, 'macro_avg': {'precision': 0.6001112901660781, 'recall': 0.4520417277842448, 'f1': 0.4141391693459015, 'support': None}, 'weighted_avg': {'precision': 0.7644486573714037, 'recall': 0.6480552070263488, 'f1': 0.6661767570381577, 'support': None}}
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831908831908832, 'agreement': 0.6285714285714286, 'direct_attack': 0.19594594594594594, 'undercutter_attack': 0.42810457516339867, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9253731343283582, 'agreement': 0.23404255319148937, 'direct_attack': 0.25663716814159293, 'undercutter_attack': 0.6931216931216931, 'partial': 0.050505050505050504}, 'f1': {'support': 0.9037900874635569, 'agreement': 0.34108527131782945, 'direct_attack': 0.22222222222222224, 'undercutter_attack': 0.5292929292929294, 'partial': 0.08620689655172413}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7145545796737767, 'recall': 0.7145545796737767, 'f1': 0.7145545796737767, 'support': None}, 'macro_avg': {'precision': 0.4859860959860959, 'recall': 0.4319359198576368, 'f1': 0.4165194813696524, 'support': None}, 'weighted_avg': {'precision': 0.7138955896078193, 'recall': 0.7145545796737767, 'f1': 0.6939241069101142, 'support': None}}
Loss: tensor(0.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8781869688385269, 'agreement': 0.5197368421052632, 'direct_attack': 0.27049180327868855, 'undercutter_attack': 0.48582995951417, 'partial': 0.42857142857142855}, 'recall': {'support': 0.9253731343283582, 'agreement': 0.42021276595744683, 'direct_attack': 0.2920353982300885, 'undercutter_attack': 0.6349206349206349, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9011627906976744, 'agreement': 0.4647058823529412, 'direct_attack': 0.28085106382978725, 'undercutter_attack': 0.5504587155963302, 'partial': 0.10619469026548672}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7327478042659975, 'recall': 0.7327478042659975, 'f1': 0.7327478042659975, 'support': None}, 'macro_avg': {'precision': 0.5165634004616153, 'recall': 0.4666295988085178, 'f1': 0.460674628548444, 'support': None}, 'weighted_avg': {'precision': 0.7183842142696049, 'recall': 0.7327478042659975, 'f1': 0.7147549889148501, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8728971962616823, 'agreement': 0.5, 'direct_attack': 0.2706766917293233, 'undercutter_attack': 0.5023041474654378, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9284294234592445, 'agreement': 0.4308510638297872, 'direct_attack': 0.3185840707964602, 'undercutter_attack': 0.5797872340425532, 'partial': 0.04040404040404041}, 'f1': {'support': 0.8998073217726397, 'agreement': 0.46285714285714286, 'direct_attack': 0.2926829268292683, 'undercutter_attack': 0.5382716049382715, 'partial': 0.07207207207207209}, 'support': {'support': 1006, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 188, 'partial': 99}, 'micro_avg': {'precision': 0.7302383939774153, 'recall': 0.7302383939774153, 'f1': 0.7302383939774153, 'support': None}, 'macro_avg': {'precision': 0.49584227375795537, 'recall': 0.45961116650641703, 'f1': 0.4531382136938789, 'support': None}, 'weighted_avg': {'precision': 0.7090051601807831, 'recall': 0.7302383939774153, 'f1': 0.7111836111390565, 'support': None}}
Loss: tensor(9.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8910505836575876, 'agreement': 0.5146198830409356, 'direct_attack': 0.2421875, 'undercutter_attack': 0.47843137254901963, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9114427860696518, 'agreement': 0.46808510638297873, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6455026455026455, 'partial': 0.04040404040404041}, 'f1': {'support': 0.9011313330054108, 'agreement': 0.49025069637883006, 'direct_attack': 0.2572614107883817, 'undercutter_attack': 0.5495495495495496, 'partial': 0.07207207207207209}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7283563362609786, 'recall': 0.7283563362609786, 'f1': 0.7283563362609787, 'support': None}, 'macro_avg': {'precision': 0.49192453451617524, 'recall': 0.46795417230903136, 'f1': 0.45405301235884893, 'support': None}, 'weighted_avg': {'precision': 0.7170922782304493, 'recall': 0.7283563362609786, 'f1': 0.7138485947357246, 'support': None}}
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8918128654970761, 'agreement': 0.5088757396449705, 'direct_attack': 0.23846153846153847, 'undercutter_attack': 0.47470817120622566, 'partial': 0.3333333333333333}, 'recall': {'support': 0.9095427435387674, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.648936170212766, 'partial': 0.04040404040404041}, 'f1': {'support': 0.9005905511811023, 'agreement': 0.48179271708683474, 'direct_attack': 0.2551440329218107, 'undercutter_attack': 0.5483146067415731, 'partial': 0.07207207207207209}, 'support': {'support': 1006, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 188, 'partial': 99}, 'micro_avg': {'precision': 0.726474278544542, 'recall': 0.726474278544542, 'f1': 0.726474278544542, 'support': None}, 'macro_avg': {'precision': 0.4894383296286288, 'recall': 0.46613320917041057, 'f1': 0.4515827960006786, 'support': None}, 'weighted_avg': {'precision': 0.716451487939923, 'recall': 0.726474278544542, 'f1': 0.7124345559744224, 'support': None}}
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8923679060665362, 'agreement': 0.5119047619047619, 'direct_attack': 0.23484848484848486, 'undercutter_attack': 0.46923076923076923, 'partial': 0.4166666666666667}, 'recall': {'support': 0.9074626865671642, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6455026455026455, 'partial': 0.050505050505050504}, 'f1': {'support': 0.8998519980266404, 'agreement': 0.48314606741573035, 'direct_attack': 0.2530612244897959, 'undercutter_attack': 0.5434298440979954, 'partial': 0.0900900900900901}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5050037177434438, 'recall': 0.46705069485426787, 'f1': 0.45391584482405045, 'support': None}, 'weighted_avg': {'precision': 0.7211670859519813, 'recall': 0.7252195734002509, 'f1': 0.7122997468705884, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8914956011730205, 'agreement': 0.5119047619047619, 'direct_attack': 0.2366412213740458, 'undercutter_attack': 0.46923076923076923, 'partial': 0.4166666666666667}, 'recall': {'support': 0.9074626865671642, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6455026455026455, 'partial': 0.050505050505050504}, 'f1': {'support': 0.8994082840236687, 'agreement': 0.48314606741573035, 'direct_attack': 0.2540983606557377, 'undercutter_attack': 0.5434298440979954, 'partial': 0.0900900900900901}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5051878040698529, 'recall': 0.46705069485426787, 'f1': 0.45403452925664445, 'support': None}, 'weighted_avg': {'precision': 0.7207441956191112, 'recall': 0.7252195734002509, 'f1': 0.7120935133785965, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8931372549019608, 'agreement': 0.5088757396449705, 'direct_attack': 0.23484848484848486, 'undercutter_attack': 0.47126436781609193, 'partial': 0.4166666666666667}, 'recall': {'support': 0.9064676616915422, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6507936507936508, 'partial': 0.050505050505050504}, 'f1': {'support': 0.8997530864197532, 'agreement': 0.48179271708683474, 'direct_attack': 0.2530612244897959, 'undercutter_attack': 0.5466666666666667, 'partial': 0.0900900900900901}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5049585027756349, 'recall': 0.4679098909373446, 'f1': 0.45427275695062813, 'support': None}, 'weighted_avg': {'precision': 0.7215360254296393, 'recall': 0.7252195734002509, 'f1': 0.7124615558032891, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8930323846908734, 'agreement': 0.5119047619047619, 'direct_attack': 0.23308270676691728, 'undercutter_attack': 0.47126436781609193, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9054726368159204, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.899209486166008, 'agreement': 0.48314606741573035, 'direct_attack': 0.2520325203252033, 'undercutter_attack': 0.5466666666666667, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5141645365434212, 'recall': 0.46973108798242214, 'f1': 0.4576395195432932, 'support': None}, 'weighted_avg': {'precision': 0.724488871346696, 'recall': 0.7252195734002509, 'f1': 0.713264624796039, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8930323846908734, 'agreement': 0.5119047619047619, 'direct_attack': 0.23308270676691728, 'undercutter_attack': 0.47126436781609193, 'partial': 0.46153846153846156}, 'recall': {'support': 0.904572564612326, 'agreement': 0.4574468085106383, 'direct_attack': 0.2767857142857143, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.8987654320987656, 'agreement': 0.48314606741573035, 'direct_attack': 0.25306122448979584, 'undercutter_attack': 0.5466666666666667, 'partial': 0.10714285714285714}, 'support': {'support': 1006, 'agreement': 188, 'direct_attack': 112, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5141645365434212, 'recall': 0.470040959761678, 'f1': 0.4577564495627631, 'support': None}, 'weighted_avg': {'precision': 0.7249028924746284, 'recall': 0.7252195734002509, 'f1': 0.7134626633409759, 'support': None}}
Loss: tensor(9.7818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8929273084479371, 'agreement': 0.5119047619047619, 'direct_attack': 0.23308270676691728, 'undercutter_attack': 0.46946564885496184, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9044776119402985, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.898665348492338, 'agreement': 0.48314606741573035, 'direct_attack': 0.2520325203252033, 'undercutter_attack': 0.5454545454545454, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7245922208281054, 'recall': 0.7245922208281054, 'f1': 0.7245922208281054, 'support': None}, 'macro_avg': {'precision': 0.513783777502608, 'recall': 0.4695320830072978, 'f1': 0.45728826776613485, 'support': None}, 'weighted_avg': {'precision': 0.724209348443431, 'recall': 0.7245922208281054, 'f1': 0.7127778310249415, 'support': None}}
Loss: tensor(7.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8929273084479371, 'agreement': 0.5119047619047619, 'direct_attack': 0.23484848484848486, 'undercutter_attack': 0.4714828897338403, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9044776119402985, 'agreement': 0.4574468085106383, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.656084656084656, 'partial': 0.06060606060606061}, 'f1': {'support': 0.898665348492338, 'agreement': 0.48314606741573035, 'direct_attack': 0.2530612244897959, 'undercutter_attack': 0.5486725663716814, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5145403812946971, 'recall': 0.47059028406549885, 'f1': 0.45813761278248055, 'support': None}, 'weighted_avg': {'precision': 0.7245737094530452, 'recall': 0.7252195734002509, 'f1': 0.7132323162971734, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8938053097345132, 'agreement': 0.5089820359281437, 'direct_attack': 0.23308270676691728, 'undercutter_attack': 0.4696969696969697, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9044776119402985, 'agreement': 0.4521276595744681, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.656084656084656, 'partial': 0.06060606060606061}, 'f1': {'support': 0.8991097922848664, 'agreement': 0.47887323943661975, 'direct_attack': 0.2520325203252033, 'undercutter_attack': 0.5474613686534217, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7245922208281054, 'recall': 0.7245922208281054, 'f1': 0.7245922208281054, 'support': None}, 'macro_avg': {'precision': 0.5134210967330011, 'recall': 0.4695264542782648, 'f1': 0.45692395556859366, 'support': None}, 'weighted_avg': {'precision': 0.7244456335428942, 'recall': 0.7245922208281054, 'f1': 0.7127920493034898, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8939096267190569, 'agreement': 0.509090909090909, 'direct_attack': 0.23134328358208955, 'undercutter_attack': 0.4696969696969697, 'partial': 0.46153846153846156}, 'recall': {'support': 0.904572564612326, 'agreement': 0.44919786096256686, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.656084656084656, 'partial': 0.06060606060606061}, 'f1': {'support': 0.8992094861660078, 'agreement': 0.47727272727272724, 'direct_attack': 0.25101214574898784, 'undercutter_attack': 0.5474613686534217, 'partial': 0.10714285714285714}, 'support': {'support': 1006, 'agreement': 187, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7245922208281054, 'recall': 0.7245922208281054, 'f1': 0.7245922208281054, 'support': None}, 'macro_avg': {'precision': 0.5131158501254973, 'recall': 0.46895948509029006, 'f1': 0.45641971699680034, 'support': None}, 'weighted_avg': {'precision': 0.7246423528790353, 'recall': 0.7245922208281054, 'f1': 0.7128585050723206, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8938053097345132, 'agreement': 0.5120481927710844, 'direct_attack': 0.22962962962962963, 'undercutter_attack': 0.4714828897338403, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9044776119402985, 'agreement': 0.4521276595744681, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.656084656084656, 'partial': 0.06060606060606061}, 'f1': {'support': 0.8991097922848664, 'agreement': 0.480225988700565, 'direct_attack': 0.25, 'undercutter_attack': 0.5486725663716814, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7245922208281054, 'recall': 0.7245922208281054, 'f1': 0.7245922208281054, 'support': None}, 'macro_avg': {'precision': 0.5137008966815058, 'recall': 0.4695264542782648, 'f1': 0.45703024089999406, 'support': None}, 'weighted_avg': {'precision': 0.7247742274305529, 'recall': 0.7245922208281054, 'f1': 0.7129511198390136, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8949950932286556, 'agreement': 0.5120481927710844, 'direct_attack': 0.23308270676691728, 'undercutter_attack': 0.4714828897338403, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9065606361829026, 'agreement': 0.45454545454545453, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.656084656084656, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9007407407407406, 'agreement': 0.481586402266289, 'direct_attack': 0.2520325203252033, 'undercutter_attack': 0.5486725663716814, 'partial': 0.10714285714285714}, 'support': {'support': 1006, 'agreement': 187, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.726474278544542, 'recall': 0.726474278544542, 'f1': 0.726474278544542, 'support': None}, 'macro_avg': {'precision': 0.5146294688077918, 'recall': 0.4704266181209829, 'f1': 0.4580350173693543, 'support': None}, 'weighted_avg': {'precision': 0.7260094074986733, 'recall': 0.726474278544542, 'f1': 0.714546910355784, 'support': None}}
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8949950932286556, 'agreement': 0.5120481927710844, 'direct_attack': 0.23134328358208955, 'undercutter_attack': 0.46946564885496184, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9065606361829026, 'agreement': 0.45454545454545453, 'direct_attack': 0.2743362831858407, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9007407407407406, 'agreement': 0.481586402266289, 'direct_attack': 0.25101214574898784, 'undercutter_attack': 0.5454545454545454, 'partial': 0.10714285714285714}, 'support': {'support': 1006, 'agreement': 187, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7258469259723965, 'recall': 0.7258469259723965, 'f1': 0.7258469259723964, 'support': None}, 'macro_avg': {'precision': 0.5138781359950506, 'recall': 0.4693684170627819, 'f1': 0.457187338270684, 'support': None}, 'weighted_avg': {'precision': 0.7256469148098442, 'recall': 0.7258469259723965, 'f1': 0.7140930155750743, 'support': None}}
Loss: tensor(6.8341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8940137389597644, 'agreement': 0.5120481927710844, 'direct_attack': 0.23703703703703705, 'undercutter_attack': 0.47126436781609193, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9064676616915422, 'agreement': 0.4521276595744681, 'direct_attack': 0.2831858407079646, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9001976284584979, 'agreement': 0.480225988700565, 'direct_attack': 0.25806451612903225, 'undercutter_attack': 0.5466666666666667, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7258469259723965, 'recall': 0.7258469259723965, 'f1': 0.7258469259723964, 'support': None}, 'macro_avg': {'precision': 0.5151803596244878, 'recall': 0.4706361746747373, 'f1': 0.4584595314195238, 'support': None}, 'weighted_avg': {'precision': 0.7254048471080685, 'recall': 0.7258469259723965, 'f1': 0.7139708504744166, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8947885939036382, 'agreement': 0.5120481927710844, 'direct_attack': 0.23357664233576642, 'undercutter_attack': 0.47126436781609193, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9054726368159204, 'agreement': 0.4521276595744681, 'direct_attack': 0.2831858407079646, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9000989119683482, 'agreement': 0.480225988700565, 'direct_attack': 0.25600000000000006, 'undercutter_attack': 0.5466666666666667, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5146432516730085, 'recall': 0.47043716969961286, 'f1': 0.4580268848956874, 'support': None}, 'weighted_avg': {'precision': 0.7256480745969955, 'recall': 0.7252195734002509, 'f1': 0.7137622554962605, 'support': None}}
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9680e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8947885939036382, 'agreement': 0.5120481927710844, 'direct_attack': 0.23357664233576642, 'undercutter_attack': 0.47126436781609193, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9054726368159204, 'agreement': 0.4521276595744681, 'direct_attack': 0.2831858407079646, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9000989119683482, 'agreement': 0.480225988700565, 'direct_attack': 0.25600000000000006, 'undercutter_attack': 0.5466666666666667, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5146432516730085, 'recall': 0.47043716969961286, 'f1': 0.4580268848956874, 'support': None}, 'weighted_avg': {'precision': 0.7256480745969955, 'recall': 0.7252195734002509, 'f1': 0.7137622554962605, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8956692913385826, 'agreement': 0.5149700598802395, 'direct_attack': 0.23357664233576642, 'undercutter_attack': 0.47126436781609193, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9054726368159204, 'agreement': 0.4574468085106383, 'direct_attack': 0.2831858407079646, 'undercutter_attack': 0.6507936507936508, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9005442850074221, 'agreement': 0.48450704225352104, 'direct_attack': 0.25600000000000006, 'undercutter_attack': 0.5466666666666667, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7258469259723965, 'recall': 0.7258469259723965, 'f1': 0.7258469259723964, 'support': None}, 'macro_avg': {'precision': 0.5154037645818285, 'recall': 0.471500999486847, 'f1': 0.45897217021409337, 'support': None}, 'weighted_avg': {'precision': 0.7265479566162177, 'recall': 0.7258469259723965, 'f1': 0.7145479756795885, 'support': None}}
Loss: tensor(6.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9736e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8956692913385826, 'agreement': 0.5119047619047619, 'direct_attack': 0.23357664233576642, 'undercutter_attack': 0.46923076923076923, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9054726368159204, 'agreement': 0.4574468085106383, 'direct_attack': 0.2831858407079646, 'undercutter_attack': 0.6455026455026455, 'partial': 0.06060606060606061}, 'f1': {'support': 0.9005442850074221, 'agreement': 0.48314606741573035, 'direct_attack': 0.25600000000000006, 'undercutter_attack': 0.5434298440979954, 'partial': 0.10714285714285714}, 'support': {'support': 1005, 'agreement': 188, 'direct_attack': 113, 'undercutter_attack': 189, 'partial': 99}, 'micro_avg': {'precision': 0.7252195734002509, 'recall': 0.7252195734002509, 'f1': 0.7252195734002509, 'support': None}, 'macro_avg': {'precision': 0.5143839852696683, 'recall': 0.47044279842864584, 'f1': 0.458052610732801, 'support': None}, 'weighted_avg': {'precision': 0.7259453053288804, 'recall': 0.7252195734002509, 'f1': 0.7140036703251448, 'support': None}}


		-------------RUN 3-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.47101449275362317, 'agreement': 0.13141524105754276, 'direct_attack': 0.2222222222222222, 'undercutter_attack': 0.09090909090909091, 'partial': 0.03260869565217391}, 'recall': {'support': 0.07471264367816093, 'agreement': 0.7544642857142857, 'direct_attack': 0.013157894736842105, 'undercutter_attack': 0.0051813471502590676, 'partial': 0.030927835051546393}, 'f1': {'support': 0.12896825396825398, 'agreement': 0.223841059602649, 'direct_attack': 0.02484472049689441, 'undercutter_attack': 0.00980392156862745, 'partial': 0.031746031746031744}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.15625, 'recall': 0.15625, 'f1': 0.15625, 'support': None}, 'macro_avg': {'precision': 0.18963394851893062, 'recall': 0.17568880126621883, 'f1': 0.08384079747649131, 'support': None}, 'weighted_avg': {'precision': 0.3214231109987207, 'recall': 0.15625, 'f1': 0.1113871730214925, 'support': None}}
Loss: tensor(0.8435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8909487459105779, 'agreement': 0.656, 'direct_attack': 0.19230769230769232, 'undercutter_attack': 0.3422982885085575, 'partial': 0.0}, 'recall': {'support': 0.9390804597701149, 'agreement': 0.36607142857142855, 'direct_attack': 0.09868421052631579, 'undercutter_attack': 0.7253886010362695, 'partial': 0.0}, 'f1': {'support': 0.9143816452154448, 'agreement': 0.4699140401146132, 'direct_attack': 0.13043478260869565, 'undercutter_attack': 0.4651162790697675, 'partial': 0.0}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.6861979166666666, 'recall': 0.6861979166666666, 'f1': 0.6861979166666666, 'support': None}, 'macro_avg': {'precision': 0.41631094534536556, 'recall': 0.42584493998082573, 'f1': 0.3959693494017042, 'support': None}, 'weighted_avg': {'precision': 0.6623461900098463, 'recall': 0.6861979166666666, 'f1': 0.6577905632422508, 'support': None}}
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3648, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8727655099894848, 'agreement': 0.686046511627907, 'direct_attack': 0.2235294117647059, 'undercutter_attack': 0.44981412639405205, 'partial': 0.13793103448275862}, 'recall': {'support': 0.9540229885057471, 'agreement': 0.26339285714285715, 'direct_attack': 0.125, 'undercutter_attack': 0.6269430051813472, 'partial': 0.20618556701030927}, 'f1': {'support': 0.9115870400878637, 'agreement': 0.3806451612903226, 'direct_attack': 0.16033755274261605, 'undercutter_attack': 0.5238095238095238, 'partial': 0.1652892561983471}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.6829427083333334, 'recall': 0.6829427083333334, 'f1': 0.6829427083333334, 'support': None}, 'macro_avg': {'precision': 0.4740173188517816, 'recall': 0.4351088835680521, 'f1': 0.4283337068257347, 'support': None}, 'weighted_avg': {'precision': 0.6817384893376418, 'recall': 0.6829427083333334, 'f1': 0.6639614876099147, 'support': None}}
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1213, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8878205128205128, 'agreement': 0.6013986013986014, 'direct_attack': 0.4, 'undercutter_attack': 0.38478747203579416, 'partial': 0.0}, 'recall': {'support': 0.9551724137931035, 'agreement': 0.38392857142857145, 'direct_attack': 0.02631578947368421, 'undercutter_attack': 0.8911917098445595, 'partial': 0.0}, 'f1': {'support': 0.9202657807308969, 'agreement': 0.46866485013623976, 'direct_attack': 0.04938271604938271, 'undercutter_attack': 0.5375, 'partial': 0.0}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7115885416666666, 'recall': 0.7115885416666666, 'f1': 0.7115885416666665, 'support': None}, 'macro_avg': {'precision': 0.4548013172509816, 'recall': 0.45132169690798374, 'f1': 0.3951626693833039, 'support': None}, 'weighted_avg': {'precision': 0.6785033300586205, 'recall': 0.7115885416666666, 'f1': 0.662015513350198, 'support': None}}
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0250, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8378378378378378, 'agreement': 0.6372549019607843, 'direct_attack': 0.3225806451612903, 'undercutter_attack': 0.43636363636363634, 'partial': 0.13178294573643412}, 'recall': {'support': 0.9620689655172414, 'agreement': 0.29017857142857145, 'direct_attack': 0.06578947368421052, 'undercutter_attack': 0.6217616580310881, 'partial': 0.17525773195876287}, 'f1': {'support': 0.8956661316211877, 'agreement': 0.39877300613496935, 'direct_attack': 0.1092896174863388, 'undercutter_attack': 0.5128205128205128, 'partial': 0.15044247787610618}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.6829427083333334, 'recall': 0.6829427083333334, 'f1': 0.6829427083333334, 'support': None}, 'macro_avg': {'precision': 0.4731639934119965, 'recall': 0.4230112801239748, 'f1': 0.4133983491878229, 'support': None}, 'weighted_avg': {'precision': 0.6625634131362413, 'recall': 0.6829427083333334, 'f1': 0.6502174408013875, 'support': None}}
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9143206854345165, 'agreement': 0.45901639344262296, 'direct_attack': 0.22277227722772278, 'undercutter_attack': 0.33047210300429186, 'partial': 0.225}, 'recall': {'support': 0.8586206896551725, 'agreement': 0.5, 'direct_attack': 0.29605263157894735, 'undercutter_attack': 0.39896373056994816, 'partial': 0.09278350515463918}, 'f1': {'support': 0.8855957320687612, 'agreement': 0.4786324786324786, 'direct_attack': 0.2542372881355932, 'undercutter_attack': 0.3615023474178404, 'partial': 0.13138686131386865}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.64453125, 'recall': 0.64453125, 'f1': 0.64453125, 'support': None}, 'macro_avg': {'precision': 0.4303162918218308, 'recall': 0.42928411139174144, 'f1': 0.42227094151370836, 'support': None}, 'weighted_avg': {'precision': 0.6625951630713666, 'recall': 0.64453125, 'f1': 0.650286789394008, 'support': None}}
Loss: tensor(0.5061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8921568627450981, 'agreement': 0.5918367346938775, 'direct_attack': 0.265625, 'undercutter_attack': 0.3962765957446808, 'partial': 0.22580645161290322}, 'recall': {'support': 0.9413793103448276, 'agreement': 0.38839285714285715, 'direct_attack': 0.1118421052631579, 'undercutter_attack': 0.772020725388601, 'partial': 0.07216494845360824}, 'f1': {'support': 0.9161073825503356, 'agreement': 0.46900269541778977, 'direct_attack': 0.1574074074074074, 'undercutter_attack': 0.523725834797891, 'partial': 0.109375}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7024739583333334, 'recall': 0.7024739583333334, 'f1': 0.7024739583333334, 'support': None}, 'macro_avg': {'precision': 0.4743403289593119, 'recall': 0.45715998931861046, 'f1': 0.4351236640346848, 'support': None}, 'weighted_avg': {'precision': 0.6819710338182546, 'recall': 0.7024739583333334, 'f1': 0.6755757901264947, 'support': None}}
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9011235955056179, 'agreement': 0.5198237885462555, 'direct_attack': 0.2966360856269113, 'undercutter_attack': 0.4588235294117647, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9218390804597701, 'agreement': 0.5267857142857143, 'direct_attack': 0.6381578947368421, 'undercutter_attack': 0.20207253886010362, 'partial': 0.020618556701030927}, 'f1': {'support': 0.9113636363636364, 'agreement': 0.5232815964523281, 'direct_attack': 0.40501043841336115, 'undercutter_attack': 0.2805755395683453, 'partial': 0.038461538461538464}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.6888020833333334, 'recall': 0.6888020833333334, 'f1': 0.6888020833333334, 'support': None}, 'macro_avg': {'precision': 0.49242425696096703, 'recall': 0.4618947570086922, 'f1': 0.43173854985184185, 'support': None}, 'weighted_avg': {'precision': 0.6912590941603488, 'recall': 0.6888020833333334, 'f1': 0.6702766121406093, 'support': None}}
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8757894736842106, 'agreement': 0.6881720430107527, 'direct_attack': 0.2864864864864865, 'undercutter_attack': 0.4444444444444444, 'partial': 0.20168067226890757}, 'recall': {'support': 0.9563218390804598, 'agreement': 0.2857142857142857, 'direct_attack': 0.34868421052631576, 'undercutter_attack': 0.43523316062176165, 'partial': 0.24742268041237114}, 'f1': {'support': 0.9142857142857143, 'agreement': 0.4037854889589905, 'direct_attack': 0.31454005934718104, 'undercutter_attack': 0.43979057591623033, 'partial': 0.22222222222222224}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.6881510416666666, 'recall': 0.6881510416666666, 'f1': 0.6881510416666666, 'support': None}, 'macro_avg': {'precision': 0.49931462397896037, 'recall': 0.45467523527103876, 'f1': 0.4589248121460677, 'support': None}, 'weighted_avg': {'precision': 0.6933425316884633, 'recall': 0.6881510416666666, 'f1': 0.6771625954970996, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8862660944206009, 'agreement': 0.5675675675675675, 'direct_attack': 0.2932330827067669, 'undercutter_attack': 0.4377358490566038, 'partial': 0.23809523809523808}, 'recall': {'support': 0.9494252873563218, 'agreement': 0.46875, 'direct_attack': 0.2565789473684211, 'undercutter_attack': 0.6010362694300518, 'partial': 0.05154639175257732}, 'f1': {'support': 0.9167591564927858, 'agreement': 0.5134474327628361, 'direct_attack': 0.2736842105263158, 'undercutter_attack': 0.5065502183406114, 'partial': 0.0847457627118644}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7102864583333334, 'recall': 0.7102864583333334, 'f1': 0.7102864583333334, 'support': None}, 'macro_avg': {'precision': 0.4845795663693554, 'recall': 0.46546737918147435, 'f1': 0.4590373561668827, 'support': None}, 'weighted_avg': {'precision': 0.6838127101664382, 'recall': 0.7102864583333334, 'f1': 0.6902195457098879, 'support': None}}
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8908108108108108, 'agreement': 0.5797872340425532, 'direct_attack': 0.30434782608695654, 'undercutter_attack': 0.44106463878326996, 'partial': 0.22727272727272727}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.48660714285714285, 'direct_attack': 0.27631578947368424, 'undercutter_attack': 0.6010362694300518, 'partial': 0.05154639175257732}, 'f1': {'support': 0.9181058495821727, 'agreement': 0.5291262135922331, 'direct_attack': 0.28965517241379307, 'undercutter_attack': 0.5087719298245613, 'partial': 0.08403361344537814}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7135416666666666, 'recall': 0.7135416666666666, 'f1': 0.7135416666666666, 'support': None}, 'macro_avg': {'precision': 0.48865664739926346, 'recall': 0.472526406059013, 'f1': 0.4659385557716277, 'support': None}, 'weighted_avg': {'precision': 0.689003610173685, 'recall': 0.7135416666666666, 'f1': 0.6950834571278574, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8815368196371398, 'agreement': 0.572289156626506, 'direct_attack': 0.30303030303030304, 'undercutter_attack': 0.42805755395683454, 'partial': 0.21739130434782608}, 'recall': {'support': 0.9494252873563218, 'agreement': 0.42410714285714285, 'direct_attack': 0.2631578947368421, 'undercutter_attack': 0.616580310880829, 'partial': 0.05154639175257732}, 'f1': {'support': 0.9142224681793026, 'agreement': 0.4871794871794872, 'direct_attack': 0.28169014084507044, 'undercutter_attack': 0.505307855626327, 'partial': 0.08333333333333333}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7063802083333334, 'recall': 0.7063802083333334, 'f1': 0.7063802083333334, 'support': None}, 'macro_avg': {'precision': 0.4804610275197219, 'recall': 0.4609634055167426, 'f1': 0.4543466570327041, 'support': None}, 'weighted_avg': {'precision': 0.6802685381931401, 'recall': 0.7063802083333334, 'f1': 0.6854989604960049, 'support': None}}
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8842443729903537, 'agreement': 0.5730994152046783, 'direct_attack': 0.3037037037037037, 'undercutter_attack': 0.427536231884058, 'partial': 0.19047619047619047}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.4375, 'direct_attack': 0.26973684210526316, 'undercutter_attack': 0.6113989637305699, 'partial': 0.041237113402061855}, 'f1': {'support': 0.9151414309484194, 'agreement': 0.4962025316455696, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.5031982942430703, 'partial': 0.06779661016949153}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.70703125, 'recall': 0.70703125, 'f1': 0.70703125, 'support': None}, 'macro_avg': {'precision': 0.47581198285179693, 'recall': 0.46162975626137204, 'f1': 0.4536106305441672, 'support': None}, 'weighted_avg': {'precision': 0.6802216925131721, 'recall': 0.70703125, 'f1': 0.6864873212354539, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8832976445396146, 'agreement': 0.5680473372781065, 'direct_attack': 0.2971014492753623, 'undercutter_attack': 0.4233576642335766, 'partial': 0.19047619047619047}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.42857142857142855, 'direct_attack': 0.26973684210526316, 'undercutter_attack': 0.6010362694300518, 'partial': 0.041237113402061855}, 'f1': {'support': 0.9146341463414634, 'agreement': 0.48854961832061067, 'direct_attack': 0.2827586206896552, 'undercutter_attack': 0.4967880085653105, 'partial': 0.06779661016949153}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7044270833333334, 'recall': 0.7044270833333334, 'f1': 0.7044270833333334, 'support': None}, 'macro_avg': {'precision': 0.47245605716057015, 'recall': 0.45777150311555415, 'f1': 0.4501054008173063, 'support': None}, 'weighted_avg': {'precision': 0.6777703087649, 'recall': 0.7044270833333334, 'f1': 0.6839859954461348, 'support': None}}
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8842443729903537, 'agreement': 0.5705882352941176, 'direct_attack': 0.3, 'undercutter_attack': 0.4264705882352941, 'partial': 0.19047619047619047}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.4330357142857143, 'direct_attack': 0.27631578947368424, 'undercutter_attack': 0.6010362694300518, 'partial': 0.041237113402061855}, 'f1': {'support': 0.9151414309484194, 'agreement': 0.49238578680203055, 'direct_attack': 0.28767123287671237, 'undercutter_attack': 0.4989247311827957, 'partial': 0.06779661016949153}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7057291666666666, 'recall': 0.7057291666666666, 'f1': 0.7057291666666666, 'support': None}, 'macro_avg': {'precision': 0.4743558773991913, 'recall': 0.45998014973209556, 'f1': 0.45238395839588985, 'support': None}, 'weighted_avg': {'precision': 0.6793550671960237, 'recall': 0.7057291666666666, 'f1': 0.6855873911919012, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8832976445396146, 'agreement': 0.5748502994011976, 'direct_attack': 0.30985915492957744, 'undercutter_attack': 0.4264705882352941, 'partial': 0.19047619047619047}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.42857142857142855, 'direct_attack': 0.2894736842105263, 'undercutter_attack': 0.6010362694300518, 'partial': 0.041237113402061855}, 'f1': {'support': 0.9146341463414634, 'agreement': 0.4910485933503836, 'direct_attack': 0.2993197278911564, 'undercutter_attack': 0.4989247311827957, 'partial': 0.06779661016949153}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7063802083333334, 'recall': 0.7063802083333334, 'f1': 0.7063802083333334, 'support': None}, 'macro_avg': {'precision': 0.4769907755163748, 'recall': 0.46171887153660685, 'f1': 0.45434476178705807, 'support': None}, 'weighted_avg': {'precision': 0.6804160308399939, 'recall': 0.7063802083333334, 'f1': 0.6862577702940983, 'support': None}}
Loss: tensor(4.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8832976445396146, 'agreement': 0.5802469135802469, 'direct_attack': 0.3076923076923077, 'undercutter_attack': 0.4233576642335766, 'partial': 0.21739130434782608}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.41964285714285715, 'direct_attack': 0.2894736842105263, 'undercutter_attack': 0.6010362694300518, 'partial': 0.05154639175257732}, 'f1': {'support': 0.9146341463414634, 'agreement': 0.48704663212435234, 'direct_attack': 0.2983050847457627, 'undercutter_attack': 0.4967880085653105, 'partial': 0.08333333333333333}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7057291666666666, 'recall': 0.7057291666666666, 'f1': 0.7057291666666666, 'support': None}, 'macro_avg': {'precision': 0.4823971668787144, 'recall': 0.4619950129209956, 'f1': 0.45602144102204456, 'support': None}, 'weighted_avg': {'precision': 0.6822971848173766, 'recall': 0.7057291666666666, 'f1': 0.6862864223832826, 'support': None}}
Loss: tensor(8.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8832976445396146, 'agreement': 0.5875, 'direct_attack': 0.30935251798561153, 'undercutter_attack': 0.4265232974910394, 'partial': 0.25}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.41964285714285715, 'direct_attack': 0.28289473684210525, 'undercutter_attack': 0.616580310880829, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9146341463414634, 'agreement': 0.48958333333333337, 'direct_attack': 0.29553264604810997, 'undercutter_attack': 0.5042372881355932, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7076822916666666, 'recall': 0.7076822916666666, 'f1': 0.7076822916666666, 'support': None}, 'macro_avg': {'precision': 0.4913346920032532, 'recall': 0.4658498874075699, 'f1': 0.46063219351550166, 'support': None}, 'weighted_avg': {'precision': 0.6859762564446928, 'recall': 0.7076822916666666, 'f1': 0.6883183382187278, 'support': None}}
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8823529411764706, 'agreement': 0.5849056603773585, 'direct_attack': 0.30935251798561153, 'undercutter_attack': 0.4265232974910394, 'partial': 0.25}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.41517857142857145, 'direct_attack': 0.28289473684210525, 'undercutter_attack': 0.616580310880829, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9141274238227146, 'agreement': 0.4856396866840731, 'direct_attack': 0.29553264604810997, 'undercutter_attack': 0.5042372881355932, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.70703125, 'recall': 0.70703125, 'f1': 0.70703125, 'support': None}, 'macro_avg': {'precision': 0.49062688340609606, 'recall': 0.46495703026471275, 'f1': 0.4597421196818998, 'support': None}, 'weighted_avg': {'precision': 0.6850628293604436, 'recall': 0.70703125, 'f1': 0.6874562122807423, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8834224598930481, 'agreement': 0.5911949685534591, 'direct_attack': 0.30935251798561153, 'undercutter_attack': 0.4265232974910394, 'partial': 0.25}, 'recall': {'support': 0.9494252873563218, 'agreement': 0.41964285714285715, 'direct_attack': 0.28289473684210525, 'undercutter_attack': 0.616580310880829, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9152354570637119, 'agreement': 0.4908616187989557, 'direct_attack': 0.29553264604810997, 'undercutter_attack': 0.5042372881355932, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.4920986487846316, 'recall': 0.4660797724650411, 'f1': 0.46100811275307585, 'support': None}, 'weighted_avg': {'precision': 0.6865858022216864, 'recall': 0.7083333333333334, 'f1': 0.6888453410004046, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8832976445396146, 'agreement': 0.5949367088607594, 'direct_attack': 0.302158273381295, 'undercutter_attack': 0.4234875444839858, 'partial': 0.25}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.41964285714285715, 'direct_attack': 0.27631578947368424, 'undercutter_attack': 0.616580310880829, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9146341463414634, 'agreement': 0.49214659685863876, 'direct_attack': 0.288659793814433, 'undercutter_attack': 0.5021097046413502, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.70703125, 'recall': 0.70703125, 'f1': 0.70703125, 'support': None}, 'macro_avg': {'precision': 0.4907760342531309, 'recall': 0.4645340979338858, 'f1': 0.4593447590749788, 'support': None}, 'weighted_avg': {'precision': 0.6859674005036723, 'recall': 0.70703125, 'f1': 0.6877446883982595, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8832976445396146, 'agreement': 0.5987261146496815, 'direct_attack': 0.302158273381295, 'undercutter_attack': 0.425531914893617, 'partial': 0.25}, 'recall': {'support': 0.9482758620689655, 'agreement': 0.41964285714285715, 'direct_attack': 0.27631578947368424, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9146341463414634, 'agreement': 0.4934383202099738, 'direct_attack': 0.288659793814433, 'undercutter_attack': 0.5052631578947369, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7076822916666666, 'recall': 0.7076822916666666, 'f1': 0.7076822916666666, 'support': None}, 'macro_avg': {'precision': 0.49194278949284165, 'recall': 0.46557036736393764, 'f1': 0.4602337943959231, 'support': None}, 'weighted_avg': {'precision': 0.6867768994527462, 'recall': 0.7076822916666666, 'f1': 0.6883292994064641, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.5987261146496815, 'direct_attack': 0.302158273381295, 'undercutter_attack': 0.42402826855123676, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41964285714285715, 'direct_attack': 0.27631578947368424, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.4934383202099738, 'direct_attack': 0.288659793814433, 'undercutter_attack': 0.5042016806722689, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.70703125, 'recall': 0.70703125, 'f1': 0.70703125, 'support': None}, 'macro_avg': {'precision': 0.49161704364227327, 'recall': 0.46534048230646635, 'f1': 0.4599011034047121, 'support': None}, 'weighted_avg': {'precision': 0.6865171169812744, 'recall': 0.70703125, 'f1': 0.6878549597370909, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6025641025641025, 'direct_attack': 0.30714285714285716, 'undercutter_attack': 0.4225352112676056, 'partial': 0.2608695652173913}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41964285714285715, 'direct_attack': 0.28289473684210525, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.4947368421052632, 'direct_attack': 0.29452054794520544, 'undercutter_attack': 0.5031446540880503, 'partial': 0.1}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7076822916666666, 'recall': 0.7076822916666666, 'f1': 0.7076822916666666, 'support': None}, 'macro_avg': {'precision': 0.49525685956422194, 'recall': 0.46665627178015046, 'f1': 0.46128684254927893, 'support': None}, 'weighted_avg': {'precision': 0.6880689098165179, 'recall': 0.7076822916666666, 'f1': 0.6885436724264947, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6064516129032258, 'direct_attack': 0.30714285714285716, 'undercutter_attack': 0.42105263157894735, 'partial': 0.2608695652173913}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41964285714285715, 'direct_attack': 0.28289473684210525, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49604221635883905, 'direct_attack': 0.29452054794520544, 'undercutter_attack': 0.502092050209205, 'partial': 0.1}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7076822916666666, 'recall': 0.7076822916666666, 'f1': 0.7076822916666666, 'support': None}, 'macro_avg': {'precision': 0.49573784569431495, 'recall': 0.46665627178015046, 'f1': 0.46133739662422507, 'support': None}, 'weighted_avg': {'precision': 0.6884495507254064, 'recall': 0.7076822916666666, 'f1': 0.6886017787313019, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6038961038961039, 'direct_attack': 0.31690140845070425, 'undercutter_attack': 0.42402826855123676, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41517857142857145, 'direct_attack': 0.29605263157894735, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49206349206349215, 'direct_attack': 0.30612244897959184, 'undercutter_attack': 0.5042016806722689, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.49559966850543963, 'recall': 0.4683949935846618, 'f1': 0.46311866880844743, 'support': None}, 'weighted_avg': {'precision': 0.6887300298209547, 'recall': 0.7083333333333334, 'f1': 0.6893825392164478, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6038961038961039, 'direct_attack': 0.3142857142857143, 'undercutter_attack': 0.4245614035087719, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41517857142857145, 'direct_attack': 0.2894736842105263, 'undercutter_attack': 0.6269430051813472, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49206349206349215, 'direct_attack': 0.3013698630136986, 'undercutter_attack': 0.5062761506276151, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.4951831566639487, 'recall': 0.4681154735410294, 'f1': 0.46258304560633795, 'support': None}, 'weighted_avg': {'precision': 0.6885381740486407, 'recall': 0.7083333333333334, 'f1': 0.6891728905410351, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5116e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3905e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9960e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6038961038961039, 'direct_attack': 0.3142857142857143, 'undercutter_attack': 0.4245614035087719, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41517857142857145, 'direct_attack': 0.2894736842105263, 'undercutter_attack': 0.6269430051813472, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49206349206349215, 'direct_attack': 0.3013698630136986, 'undercutter_attack': 0.5062761506276151, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.4951831566639487, 'recall': 0.4681154735410294, 'f1': 0.46258304560633795, 'support': None}, 'weighted_avg': {'precision': 0.6885381740486407, 'recall': 0.7083333333333334, 'f1': 0.6891728905410351, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8119e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7546e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2693e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4518e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7841e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6327e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6630e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1179e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6932e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2685e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5503e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6038961038961039, 'direct_attack': 0.31690140845070425, 'undercutter_attack': 0.42402826855123676, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41517857142857145, 'direct_attack': 0.29605263157894735, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49206349206349215, 'direct_attack': 0.30612244897959184, 'undercutter_attack': 0.5042016806722689, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.49559966850543963, 'recall': 0.4683949935846618, 'f1': 0.46311866880844743, 'support': None}, 'weighted_avg': {'precision': 0.6887300298209547, 'recall': 0.7083333333333334, 'f1': 0.6893825392164478, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6638e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0271e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9666e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9658e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1171e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4207e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2988e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6008e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6597e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5721e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8144e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2358e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6038961038961039, 'direct_attack': 0.31690140845070425, 'undercutter_attack': 0.42402826855123676, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41517857142857145, 'direct_attack': 0.29605263157894735, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49206349206349215, 'direct_attack': 0.30612244897959184, 'undercutter_attack': 0.5042016806722689, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.49559966850543963, 'recall': 0.4683949935846618, 'f1': 0.46311866880844743, 'support': None}, 'weighted_avg': {'precision': 0.6887300298209547, 'recall': 0.7083333333333334, 'f1': 0.6893825392164478, 'support': None}}
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3299e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3004e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9060e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8749e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5729e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1466e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7243e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7227e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8430e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8438e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1482e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7530e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6940e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2088e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3594e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9355e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8446e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0566e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8128e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1163e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5419e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0558e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1155e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4821e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5713e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0574e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2080e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2366e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2996e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6924e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7538e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5108e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8757e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2096e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9178e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8831725616291533, 'agreement': 0.6038961038961039, 'direct_attack': 0.31690140845070425, 'undercutter_attack': 0.42402826855123676, 'partial': 0.25}, 'recall': {'support': 0.9471264367816092, 'agreement': 0.41517857142857145, 'direct_attack': 0.29605263157894735, 'undercutter_attack': 0.6217616580310881, 'partial': 0.061855670103092786}, 'f1': {'support': 0.9140321686078758, 'agreement': 0.49206349206349215, 'direct_attack': 0.30612244897959184, 'undercutter_attack': 0.5042016806722689, 'partial': 0.09917355371900827}, 'support': {'support': 870, 'agreement': 224, 'direct_attack': 152, 'undercutter_attack': 193, 'partial': 97}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.49559966850543963, 'recall': 0.4683949935846618, 'f1': 0.46311866880844743, 'support': None}, 'weighted_avg': {'precision': 0.6887300298209547, 'recall': 0.7083333333333334, 'f1': 0.6893825392164478, 'support': None}}


		-------------RUN 4-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.596039603960396, 'agreement': 0.0, 'direct_attack': 0.10606060606060606, 'undercutter_attack': 0.0, 'partial': 0.0}, 'recall': {'support': 0.7395577395577395, 'agreement': 0.0, 'direct_attack': 0.26515151515151514, 'undercutter_attack': 0.0, 'partial': 0.0}, 'f1': {'support': 0.6600877192982456, 'agreement': 0.0, 'direct_attack': 0.15151515151515152, 'undercutter_attack': 0.0, 'partial': 0.0}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.4739583333333333, 'recall': 0.4739583333333333, 'f1': 0.4739583333333333, 'support': None}, 'macro_avg': {'precision': 0.1404200420042004, 'recall': 0.20094185094185094, 'f1': 0.16232057416267942, 'support': None}, 'weighted_avg': {'precision': 0.3714108910891089, 'recall': 0.4739583333333333, 'f1': 0.4146662228487886, 'support': None}}
Loss: tensor(0.6322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5150, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.7325428194993412, 'agreement': 0.5833333333333334, 'direct_attack': 0.0, 'undercutter_attack': 0.24242424242424243, 'partial': 0.11904761904761904}, 'recall': {'support': 0.683046683046683, 'agreement': 0.2014388489208633, 'direct_attack': 0.0, 'undercutter_attack': 0.7643312101910829, 'partial': 0.049019607843137254}, 'f1': {'support': 0.7069294342021615, 'agreement': 0.29946524064171126, 'direct_attack': 0.0, 'undercutter_attack': 0.3680981595092025, 'partial': 0.06944444444444445}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.5275297619047619, 'recall': 0.5275297619047619, 'f1': 0.5275297619047619, 'support': None}, 'macro_avg': {'precision': 0.3354696028609072, 'recall': 0.3395672700003533, 'f1': 0.28878745575950393, 'support': None}, 'weighted_avg': {'precision': 0.5413516753045092, 'recall': 0.5275297619047619, 'f1': 0.5073965567455622, 'support': None}}
Loss: tensor(0.6769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8392, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8714953271028038, 'agreement': 0.38961038961038963, 'direct_attack': 0.0, 'undercutter_attack': 0.4271523178807947, 'partial': 0.2037037037037037}, 'recall': {'support': 0.9164619164619164, 'agreement': 0.2158273381294964, 'direct_attack': 0.0, 'undercutter_attack': 0.821656050955414, 'partial': 0.21568627450980393}, 'f1': {'support': 0.8934131736526946, 'agreement': 0.27777777777777773, 'direct_attack': 0.0, 'undercutter_attack': 0.5620915032679739, 'partial': 0.2095238095238095}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.6897321428571429, 'recall': 0.6897321428571429, 'f1': 0.6897321428571429, 'support': None}, 'macro_avg': {'precision': 0.3783923476595384, 'recall': 0.4339263160113262, 'f1': 0.38856125284445114, 'support': None}, 'weighted_avg': {'precision': 0.6334774792429977, 'recall': 0.6897321428571429, 'f1': 0.6513907954232924, 'support': None}}
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8696655132641292, 'agreement': 0.375, 'direct_attack': 0.32432432432432434, 'undercutter_attack': 0.4782608695652174, 'partial': 0.17647058823529413}, 'recall': {'support': 0.9262899262899262, 'agreement': 0.4748201438848921, 'direct_attack': 0.09090909090909091, 'undercutter_attack': 0.7006369426751592, 'partial': 0.058823529411764705}, 'f1': {'support': 0.8970850684116597, 'agreement': 0.4190476190476191, 'direct_attack': 0.1420118343195266, 'undercutter_attack': 0.5684754521963824, 'partial': 0.08823529411764706}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7053571428571429, 'recall': 0.7053571428571429, 'f1': 0.7053571428571429, 'support': None}, 'macro_avg': {'precision': 0.444744259077793, 'recall': 0.45029592663416657, 'f1': 0.422971053618567, 'support': None}, 'weighted_avg': {'precision': 0.666614951733297, 'recall': 0.7053571428571429, 'f1': 0.6737135958777676, 'support': None}}
Loss: tensor(0.3890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1839, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8950, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8132337246531484, 'agreement': 0.4069767441860465, 'direct_attack': 0.29411764705882354, 'undercutter_attack': 0.42028985507246375, 'partial': 0.45454545454545453}, 'recall': {'support': 0.9361179361179361, 'agreement': 0.5035971223021583, 'direct_attack': 0.03787878787878788, 'undercutter_attack': 0.554140127388535, 'partial': 0.049019607843137254}, 'f1': {'support': 0.8703597944031982, 'agreement': 0.45016077170418, 'direct_attack': 0.06711409395973154, 'undercutter_attack': 0.478021978021978, 'partial': 0.08849557522123894}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.6912202380952381, 'recall': 0.6912202380952381, 'f1': 0.6912202380952381, 'support': None}, 'macro_avg': {'precision': 0.4778326851031873, 'recall': 0.4161507163061109, 'f1': 0.39083044266206535, 'support': None}, 'weighted_avg': {'precision': 0.6471091460798372, 'recall': 0.6912202380952381, 'f1': 0.6428424698926978, 'support': None}}
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0764, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5480, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8788221970554927, 'agreement': 0.39325842696629215, 'direct_attack': 0.4050632911392405, 'undercutter_attack': 0.44534412955465585, 'partial': 0.30434782608695654}, 'recall': {'support': 0.9533169533169533, 'agreement': 0.2517985611510791, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7006369426751592, 'partial': 0.13725490196078433}, 'f1': {'support': 0.9145550972304066, 'agreement': 0.30701754385964913, 'direct_attack': 0.3033175355450237, 'undercutter_attack': 0.5445544554455446, 'partial': 0.18918918918918923}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7194940476190477, 'recall': 0.7194940476190477, 'f1': 0.7194940476190476, 'support': None}, 'macro_avg': {'precision': 0.48536717416052755, 'recall': 0.45708632030564367, 'f1': 0.45172676425396263, 'support': None}, 'weighted_avg': {'precision': 0.6878385794515001, 'recall': 0.7194940476190477, 'f1': 0.6934178193721974, 'support': None}}
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0899, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0909, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9179487179487179, 'agreement': 0.33191489361702126, 'direct_attack': 0.28104575163398693, 'undercutter_attack': 0.43870967741935485, 'partial': 0.23809523809523808}, 'recall': {'support': 0.8796068796068796, 'agreement': 0.5611510791366906, 'direct_attack': 0.32575757575757575, 'undercutter_attack': 0.43312101910828027, 'partial': 0.049019607843137254}, 'f1': {'support': 0.8983688833124216, 'agreement': 0.41711229946524064, 'direct_attack': 0.3017543859649122, 'undercutter_attack': 0.4358974358974359, 'partial': 0.08130081300813008}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.6770833333333334, 'recall': 0.6770833333333334, 'f1': 0.6770833333333334, 'support': None}, 'macro_avg': {'precision': 0.4415428557428638, 'recall': 0.44973123229051265, 'f1': 0.42688676352962796, 'support': None}, 'weighted_avg': {'precision': 0.6872080353268316, 'recall': 0.6770833333333334, 'f1': 0.6739665475833889, 'support': None}}
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9142857142857143, 'agreement': 0.3300970873786408, 'direct_attack': 0.2561576354679803, 'undercutter_attack': 0.40853658536585363, 'partial': 1.0}, 'recall': {'support': 0.8648648648648649, 'agreement': 0.4892086330935252, 'direct_attack': 0.3939393939393939, 'undercutter_attack': 0.4267515923566879, 'partial': 0.00980392156862745}, 'f1': {'support': 0.888888888888889, 'agreement': 0.3942028985507247, 'direct_attack': 0.31044776119402984, 'undercutter_attack': 0.4174454828660436, 'partial': 0.01941747572815534}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.6636904761904762, 'recall': 0.6636904761904762, 'f1': 0.6636904761904762, 'support': None}, 'macro_avg': {'precision': 0.5818154044996378, 'recall': 0.4369136811646199, 'f1': 0.4060805014455685, 'support': None}, 'weighted_avg': {'precision': 0.7366555940166778, 'recall': 0.6636904761904762, 'f1': 0.6598574302574101, 'support': None}}
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9145299145299145, 'agreement': 0.45161290322580644, 'direct_attack': 0.3076923076923077, 'undercutter_attack': 0.4372623574144487, 'partial': 0.2328767123287671}, 'recall': {'support': 0.9201474201474201, 'agreement': 0.4028776978417266, 'direct_attack': 0.15151515151515152, 'undercutter_attack': 0.732484076433121, 'partial': 0.16666666666666666}, 'f1': {'support': 0.9173300673606859, 'agreement': 0.42585551330798477, 'direct_attack': 0.20304568527918784, 'undercutter_attack': 0.5476190476190477, 'partial': 0.19428571428571428}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7120535714285714, 'recall': 0.7120535714285714, 'f1': 0.7120535714285713, 'support': None}, 'macro_avg': {'precision': 0.46879483903824887, 'recall': 0.47473820252081717, 'f1': 0.45762720557052405, 'support': None}, 'weighted_avg': {'precision': 0.699568856668694, 'recall': 0.7120535714285714, 'f1': 0.6982856807824362, 'support': None}}
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0933, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9200507614213198, 'agreement': 0.358695652173913, 'direct_attack': 0.33980582524271846, 'undercutter_attack': 0.47280334728033474, 'partial': 0.43333333333333335}, 'recall': {'support': 0.8906633906633906, 'agreement': 0.4748201438848921, 'direct_attack': 0.26515151515151514, 'undercutter_attack': 0.7197452229299363, 'partial': 0.12745098039215685}, 'f1': {'support': 0.9051186017478151, 'agreement': 0.4086687306501548, 'direct_attack': 0.2978723404255319, 'undercutter_attack': 0.5707070707070707, 'partial': 0.19696969696969696}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7083333333333334, 'recall': 0.7083333333333334, 'f1': 0.7083333333333334, 'support': None}, 'macro_avg': {'precision': 0.5049377838903238, 'recall': 0.4955662506043782, 'f1': 0.4758672881000538, 'support': None}, 'weighted_avg': {'precision': 0.7158218079644194, 'recall': 0.7083333333333334, 'f1': 0.7013263121362965, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9039145907473309, 'agreement': 0.48031496062992124, 'direct_attack': 0.3448275862068966, 'undercutter_attack': 0.46511627906976744, 'partial': 0.3793103448275862}, 'recall': {'support': 0.9361179361179361, 'agreement': 0.43884892086330934, 'direct_attack': 0.22727272727272727, 'undercutter_attack': 0.7643312101910829, 'partial': 0.10784313725490197}, 'f1': {'support': 0.9197344598672299, 'agreement': 0.45864661654135336, 'direct_attack': 0.273972602739726, 'undercutter_attack': 0.5783132530120483, 'partial': 0.16793893129770993}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5146967522963005, 'recall': 0.49488278633999155, 'f1': 0.4797211726916135, 'support': None}, 'weighted_avg': {'precision': 0.7141223279475923, 'recall': 0.7321428571428571, 'f1': 0.7116853164494605, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9027283511269276, 'agreement': 0.47619047619047616, 'direct_attack': 0.34782608695652173, 'undercutter_attack': 0.46825396825396826, 'partial': 0.3870967741935484}, 'recall': {'support': 0.9348894348894349, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7515923566878981, 'partial': 0.11764705882352941}, 'f1': {'support': 0.9185274592637297, 'agreement': 0.45283018867924524, 'direct_attack': 0.28571428571428575, 'undercutter_attack': 0.5770171149144254, 'partial': 0.18045112781954886}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7313988095238095, 'recall': 0.7313988095238095, 'f1': 0.7313988095238095, 'support': None}, 'macro_avg': {'precision': 0.5164191313442885, 'recall': 0.4956415538168195, 'f1': 0.482908035278247, 'support': None}, 'weighted_avg': {'precision': 0.7142292719268386, 'recall': 0.7313988095238095, 'f1': 0.7123041189438508, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4814e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9027283511269276, 'agreement': 0.47619047619047616, 'direct_attack': 0.3404255319148936, 'undercutter_attack': 0.468, 'partial': 0.3870967741935484}, 'recall': {'support': 0.9348894348894349, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7452229299363057, 'partial': 0.11764705882352941}, 'f1': {'support': 0.9185274592637297, 'agreement': 0.45283018867924524, 'direct_attack': 0.2831858407079646, 'undercutter_attack': 0.5749385749385749, 'partial': 0.18045112781954886}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7306547619047619, 'recall': 0.7306547619047619, 'f1': 0.7306547619047619, 'support': None}, 'macro_avg': {'precision': 0.5148882266851692, 'recall': 0.494367668466501, 'f1': 0.48198663828181265, 'support': None}, 'weighted_avg': {'precision': 0.7134727642770113, 'recall': 0.7306547619047619, 'f1': 0.7118129838865271, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9028436018957346, 'agreement': 0.48, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.46586345381526106, 'partial': 0.4}, 'recall': {'support': 0.9361179361179361, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.9191797346200242, 'agreement': 0.45454545454545453, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5714285714285715, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7306547619047619, 'recall': 0.7306547619047619, 'f1': 0.7306547619047619, 'support': None}, 'macro_avg': {'precision': 0.5164080778088658, 'recall': 0.4933394833618828, 'f1': 0.48153473935963936, 'support': None}, 'weighted_avg': {'precision': 0.7139696831786637, 'recall': 0.7306547619047619, 'f1': 0.7118351889889922, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9028436018957346, 'agreement': 0.48, 'direct_attack': 0.3368421052631579, 'undercutter_attack': 0.468, 'partial': 0.4}, 'recall': {'support': 0.9361179361179361, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7452229299363057, 'partial': 0.11764705882352941}, 'f1': {'support': 0.9191797346200242, 'agreement': 0.45454545454545453, 'direct_attack': 0.28193832599118945, 'undercutter_attack': 0.5749385749385749, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7313988095238095, 'recall': 0.7313988095238095, 'f1': 0.7313988095238095, 'support': None}, 'macro_avg': {'precision': 0.5175371414317784, 'recall': 0.4946133687122013, 'f1': 0.48248405438268493, 'support': None}, 'weighted_avg': {'precision': 0.7145638763674589, 'recall': 0.7313988095238095, 'f1': 0.7123666607173851, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4838709677419355, 'direct_attack': 0.3368421052631579, 'undercutter_attack': 0.468, 'partial': 0.4}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7452229299363057, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.45627376425855515, 'direct_attack': 0.28193832599118945, 'undercutter_attack': 0.5749385749385749, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.51833433057735, 'recall': 0.49485906895790155, 'f1': 0.48296001412703865, 'support': None}, 'weighted_avg': {'precision': 0.7150338589542667, 'recall': 0.7321428571428571, 'f1': 0.7129399844548676, 'support': None}}
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4878048780487805, 'direct_attack': 0.3368421052631579, 'undercutter_attack': 0.46613545816733065, 'partial': 0.4}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7452229299363057, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.4580152671755725, 'direct_attack': 0.28193832599118945, 'undercutter_attack': 0.5735294117647058, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5187482042721852, 'recall': 0.49485906895790155, 'f1': 0.4830264820756683, 'support': None}, 'weighted_avg': {'precision': 0.7152229069192387, 'recall': 0.7321428571428571, 'f1': 0.712955483180439, 'support': None}}
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4918032786885246, 'direct_attack': 0.3402061855670103, 'undercutter_attack': 0.464, 'partial': 0.4}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.4316546762589928, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.45977011494252873, 'direct_attack': 0.28820960698689957, 'undercutter_attack': 0.5700245700245701, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5197936088274384, 'recall': 0.4951003351227346, 'f1': 0.4839307394801744, 'support': None}, 'weighted_avg': {'precision': 0.7157173781668296, 'recall': 0.7321428571428571, 'f1': 0.7133434837591887, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4878048780487805, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.464, 'partial': 0.4}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.4316546762589928, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.4580152671755725, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5700245700245701, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7313988095238095, 'recall': 0.7313988095238095, 'f1': 0.7313988095238095, 'support': None}, 'macro_avg': {'precision': 0.5176193582527542, 'recall': 0.493585183607583, 'f1': 0.48207819940659624, 'support': None}, 'weighted_avg': {'precision': 0.714628840827715, 'recall': 0.7313988095238095, 'f1': 0.7124246144266512, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4918032786885246, 'direct_attack': 0.3402061855670103, 'undercutter_attack': 0.464, 'partial': 0.4}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.4316546762589928, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.45977011494252873, 'direct_attack': 0.28820960698689957, 'undercutter_attack': 0.5700245700245701, 'partial': 0.1818181818181818}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5197936088274384, 'recall': 0.4951003351227346, 'f1': 0.4839307394801744, 'support': None}, 'weighted_avg': {'precision': 0.7157173781668296, 'recall': 0.7321428571428571, 'f1': 0.7133434837591887, 'support': None}}
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4959349593495935, 'direct_attack': 0.3402061855670103, 'undercutter_attack': 0.464, 'partial': 0.41379310344827586}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46564885496183206, 'direct_attack': 0.28820960698689957, 'undercutter_attack': 0.5700245700245701, 'partial': 0.183206106870229}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7328869047619048, 'recall': 0.7328869047619048, 'f1': 0.7328869047619049, 'support': None}, 'macro_avg': {'precision': 0.5233785656493073, 'recall': 0.49653918404359787, 'f1': 0.4853840724944445, 'support': None}, 'weighted_avg': {'precision': 0.7171914854314224, 'recall': 0.7328869047619048, 'f1': 0.714056812046385, 'support': None}}
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.49193548387096775, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.464, 'partial': 0.41379310344827586}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46387832699619774, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5700245700245701, 'partial': 0.183206106870229}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5212041001068467, 'recall': 0.49502403252844634, 'f1': 0.48352839638113077, 'support': None}, 'weighted_avg': {'precision': 0.7161028369296558, 'recall': 0.7321428571428571, 'f1': 0.7131363210266332, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6358e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.49193548387096775, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.464, 'partial': 0.41379310344827586}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46387832699619774, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5700245700245701, 'partial': 0.183206106870229}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5212041001068467, 'recall': 0.49502403252844634, 'f1': 0.48352839638113077, 'support': None}, 'weighted_avg': {'precision': 0.7161028369296558, 'recall': 0.7321428571428571, 'f1': 0.7131363210266332, 'support': None}}
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3230e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.49193548387096775, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.464, 'partial': 0.41379310344827586}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46387832699619774, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5700245700245701, 'partial': 0.183206106870229}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5212041001068467, 'recall': 0.49502403252844634, 'f1': 0.48352839638113077, 'support': None}, 'weighted_avg': {'precision': 0.7161028369296558, 'recall': 0.7321428571428571, 'f1': 0.7131363210266332, 'support': None}}
Loss: tensor(5.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8634e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.49193548387096775, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.464, 'partial': 0.41379310344827586}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46387832699619774, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5700245700245701, 'partial': 0.183206106870229}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5212041001068467, 'recall': 0.49502403252844634, 'f1': 0.48352839638113077, 'support': None}, 'weighted_avg': {'precision': 0.7161028369296558, 'recall': 0.7321428571428571, 'f1': 0.7131363210266332, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2413e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9030732860520094, 'agreement': 0.4959349593495935, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.46215139442231074, 'partial': 0.42857142857142855}, 'recall': {'support': 0.9385749385749386, 'agreement': 0.43884892086330934, 'direct_attack': 0.24242424242424243, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.9204819277108434, 'agreement': 0.46564885496183206, 'direct_attack': 0.2807017543859649, 'undercutter_attack': 0.5686274509803921, 'partial': 0.18461538461538463}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7328869047619048, 'recall': 0.7328869047619048, 'f1': 0.7328869047619049, 'support': None}, 'macro_avg': {'precision': 0.524612880345735, 'recall': 0.49526973277414665, 'f1': 0.48401507453088344, 'support': None}, 'weighted_avg': {'precision': 0.7174915690733018, 'recall': 0.7328869047619048, 'f1': 0.7136572846800292, 'support': None}}
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9180e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5395e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5387e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9030732860520094, 'agreement': 0.5, 'direct_attack': 0.3402061855670103, 'undercutter_attack': 0.46215139442231074, 'partial': 0.42857142857142855}, 'recall': {'support': 0.9385749385749386, 'agreement': 0.43884892086330934, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.9204819277108434, 'agreement': 0.4674329501915709, 'direct_attack': 0.28820960698689957, 'undercutter_attack': 0.5686274509803921, 'partial': 0.18461538461538463}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7336309523809523, 'recall': 0.7336309523809523, 'f1': 0.7336309523809524, 'support': None}, 'macro_avg': {'precision': 0.5268004589225518, 'recall': 0.4967848842892981, 'f1': 0.4858734640970182, 'support': None}, 'weighted_avg': {'precision': 0.7185869984968523, 'recall': 0.7336309523809523, 'f1': 0.7145791788617681, 'support': None}}
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0853e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3722e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4959349593495935, 'direct_attack': 0.3402061855670103, 'undercutter_attack': 0.4603174603174603, 'partial': 0.4444444444444444}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46564885496183206, 'direct_attack': 0.28820960698689957, 'undercutter_attack': 0.5672371638141809, 'partial': 0.18604651162790695}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7328869047619048, 'recall': 0.7328869047619048, 'f1': 0.7328869047619049, 'support': None}, 'macro_avg': {'precision': 0.528772325912033, 'recall': 0.49653918404359787, 'f1': 0.4853946722039023, 'support': None}, 'weighted_avg': {'precision': 0.7190875256482755, 'recall': 0.7328869047619048, 'f1': 0.7139467662950845, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0845e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4805e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2064e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2677e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1745e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5312e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4918032786885246, 'direct_attack': 0.336734693877551, 'undercutter_attack': 0.4603174603174603, 'partial': 0.4444444444444444}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.4316546762589928, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.45977011494252873, 'direct_attack': 0.2869565217391304, 'undercutter_attack': 0.5672371638141809, 'partial': 0.18604651162790695}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7321428571428571, 'recall': 0.7321428571428571, 'f1': 0.7321428571428571, 'support': None}, 'macro_avg': {'precision': 0.5272516914419274, 'recall': 0.4951003351227346, 'f1': 0.4839683071504878, 'support': None}, 'weighted_avg': {'precision': 0.718319266336596, 'recall': 0.7321428571428571, 'f1': 0.7132157007330393, 'support': None}}
Loss: tensor(5.1777e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6892e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6908e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6917e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9036e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9314e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4940e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5721e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1442e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7235e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0869e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1474e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3889e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5403e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1147e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4959349593495935, 'direct_attack': 0.336734693877551, 'undercutter_attack': 0.4603174603174603, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46564885496183206, 'direct_attack': 0.2869565217391304, 'undercutter_attack': 0.5672371638141809, 'partial': 0.1875}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7328869047619048, 'recall': 0.7328869047619048, 'f1': 0.7328869047619049, 'support': None}, 'macro_avg': {'precision': 0.5314968309929446, 'recall': 0.49653918404359787, 'f1': 0.485434752828767, 'support': None}, 'weighted_avg': {'precision': 0.720043889368946, 'recall': 0.7328869047619048, 'f1': 0.7139340048079178, 'support': None}}
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6622e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5411e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4502e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9634e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3267e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5100e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7498e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5394e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9052e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9642e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2972e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3259e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8136e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3602e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3872e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5075e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2072e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1148e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0550e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9650e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7874e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9626e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6326e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1458e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9936e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9044e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9347e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2383e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4199e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2669e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7833e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7809e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3897e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2359e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6925e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6606e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7817e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4526e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5992e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1164e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8072e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8741e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8120e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9029585798816568, 'agreement': 0.4959349593495935, 'direct_attack': 0.336734693877551, 'undercutter_attack': 0.4603174603174603, 'partial': 0.46153846153846156}, 'recall': {'support': 0.9373464373464373, 'agreement': 0.43884892086330934, 'direct_attack': 0.25, 'undercutter_attack': 0.7388535031847133, 'partial': 0.11764705882352941}, 'f1': {'support': 0.919831223628692, 'agreement': 0.46564885496183206, 'direct_attack': 0.2869565217391304, 'undercutter_attack': 0.5672371638141809, 'partial': 0.1875}, 'support': {'support': 814, 'agreement': 139, 'direct_attack': 132, 'undercutter_attack': 157, 'partial': 102}, 'micro_avg': {'precision': 0.7328869047619048, 'recall': 0.7328869047619048, 'f1': 0.7328869047619049, 'support': None}, 'macro_avg': {'precision': 0.5314968309929446, 'recall': 0.49653918404359787, 'f1': 0.485434752828767, 'support': None}, 'weighted_avg': {'precision': 0.720043889368946, 'recall': 0.7328869047619048, 'f1': 0.7139340048079178, 'support': None}}


		-------------RUN 5-----------
			------------EPOCH 1---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.06666666666666667, 'undercutter_attack': 0.0, 'partial': 0.07654145995747696}, 'recall': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.05982905982905983, 'undercutter_attack': 0.0, 'partial': 0.972972972972973}, 'f1': {'support': 0.0, 'agreement': 0.0, 'direct_attack': 0.06306306306306307, 'undercutter_attack': 0.0, 'partial': 0.14191852825229961}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.0758575197889182, 'recall': 0.0758575197889182, 'f1': 0.0758575197889182, 'support': None}, 'macro_avg': {'precision': 0.028641625324828723, 'recall': 0.20656040656040658, 'f1': 0.04099631826307254, 'support': None}, 'weighted_avg': {'precision': 0.01074940768817938, 'recall': 0.0758575197889182, 'f1': 0.015258136553023507, 'support': None}}
Loss: tensor(0.9103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6835, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4820, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4862, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6991, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5986, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4354, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 2---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9175946547884187, 'agreement': 0.4867924528301887, 'direct_attack': 0.35, 'undercutter_attack': 0.4066265060240964, 'partial': 0.0}, 'recall': {'support': 0.8738069989395546, 'agreement': 0.675392670157068, 'direct_attack': 0.05982905982905983, 'undercutter_attack': 0.8766233766233766, 'partial': 0.0}, 'f1': {'support': 0.8951656708310701, 'agreement': 0.5657894736842104, 'direct_attack': 0.10218978102189781, 'undercutter_attack': 0.5555555555555556, 'partial': 0.0}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7222955145118733, 'recall': 0.7222955145118733, 'f1': 0.7222955145118733, 'support': None}, 'macro_avg': {'precision': 0.43220272272854077, 'recall': 0.49713042110981187, 'f1': 0.4237400962185468, 'support': None}, 'weighted_avg': {'precision': 0.7004218996594695, 'recall': 0.7222955145118733, 'f1': 0.6924266339066629, 'support': None}}
Loss: tensor(0.3893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0597, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8674, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5563, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0893, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9446, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4804, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0993, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5873, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6530, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1655, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5652, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1628, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1952, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5778, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8834, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9846, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0606, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2941, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6659, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 3---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9178532311062432, 'agreement': 0.5394736842105263, 'direct_attack': 0.27647058823529413, 'undercutter_attack': 0.6, 'partial': 0.23}, 'recall': {'support': 0.8886532343584306, 'agreement': 0.643979057591623, 'direct_attack': 0.4017094017094017, 'undercutter_attack': 0.4090909090909091, 'partial': 0.2072072072072072}, 'f1': {'support': 0.9030172413793104, 'agreement': 0.5871121718377087, 'direct_attack': 0.3275261324041812, 'undercutter_attack': 0.4864864864864865, 'partial': 0.21800947867298578}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7216358839050132, 'recall': 0.7216358839050132, 'f1': 0.7216358839050132, 'support': None}, 'macro_avg': {'precision': 0.5127595007104128, 'recall': 0.5101279619915143, 'f1': 0.5044303021561345, 'support': None}, 'weighted_avg': {'precision': 0.7380291091298993, 'recall': 0.7216358839050132, 'f1': 0.7263339129186025, 'support': None}}
Loss: tensor(1.8559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1994, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0319, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1930, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8600, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6718, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0814, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0676, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7882, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0603, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3517, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3654, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3488, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4908, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8963, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0312, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1829, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3943, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1929, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0996, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7797, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2890, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7316, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0663, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0726, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4648, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1539, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9889, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5299, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3715, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4535, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 4---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9221871713985279, 'agreement': 0.6178010471204188, 'direct_attack': 0.2388888888888889, 'undercutter_attack': 0.5225806451612903, 'partial': 0.2564102564102564}, 'recall': {'support': 0.9300106044538706, 'agreement': 0.6178010471204188, 'direct_attack': 0.36752136752136755, 'undercutter_attack': 0.525974025974026, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9260823653643082, 'agreement': 0.6178010471204188, 'direct_attack': 0.2895622895622896, 'undercutter_attack': 0.5242718446601942, 'partial': 0.13333333333333333}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7447229551451188, 'recall': 0.7447229551451188, 'f1': 0.7447229551451188, 'support': None}, 'macro_avg': {'precision': 0.5115736017958764, 'recall': 0.5062794270319546, 'f1': 0.4982101760081088, 'support': None}, 'weighted_avg': {'precision': 0.7417621770746631, 'recall': 0.7447229551451188, 'f1': 0.7392561494030345, 'support': None}}
Loss: tensor(0.3412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5707, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1812, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7383, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6390, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3968, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5567, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3561, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6765, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0875, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0855, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1690, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0701, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9793, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2268, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9555, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2650, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0484, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1801, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1441, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2957, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0188, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0780, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0580, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3525, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2844, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0705, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0564, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0700, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6653, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2892, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0171, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0570, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5339, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4336, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0659, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3951, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7865, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5782, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5618, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9809, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4763, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1886, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1872, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1746, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9343, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0962, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0851, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1982, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7270, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3822, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0485, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7550, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0915, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7405, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2920, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3504, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 5---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9157245632065776, 'agreement': 0.6666666666666666, 'direct_attack': 0.24615384615384617, 'undercutter_attack': 0.5, 'partial': 0.2345679012345679}, 'recall': {'support': 0.9448568398727466, 'agreement': 0.5968586387434555, 'direct_attack': 0.41025641025641024, 'undercutter_attack': 0.3116883116883117, 'partial': 0.17117117117117117}, 'f1': {'support': 0.930062630480167, 'agreement': 0.6298342541436464, 'direct_attack': 0.3076923076923077, 'undercutter_attack': 0.384, 'partial': 0.19791666666666666}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7387862796833773, 'recall': 0.7387862796833773, 'f1': 0.7387862796833774, 'support': None}, 'macro_avg': {'precision': 0.5126225954523317, 'recall': 0.48696627434641904, 'f1': 0.48990117179655757, 'support': None}, 'weighted_avg': {'precision': 0.740566380919639, 'recall': 0.7387862796833773, 'f1': 0.7351267500555633, 'support': None}}
Loss: tensor(0.0635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0431, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0990, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8621, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0627, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0207, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0551, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0411, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7787, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0589, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3518, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9382, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0554, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0931, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1581, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4944, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2845, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5528, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0926, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0559, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5837, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0394, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3751, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8693, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0734, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0699, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0513, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1884, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3374, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0704, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1733, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0838, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1750, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1879, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0828, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0462, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4974, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1816, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2766, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2515, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0607, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0813, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2870, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7826, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2906, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1947, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0505, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4927, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6961, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0960, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0529, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2335, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7773, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2860, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5998, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0473, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1583, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0373, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2492, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4732, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0566, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3874, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0602, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9959, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0471, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5211, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3516, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3632, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0660, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0977, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7967, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8866, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6430, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2824, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0326, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6901, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0490, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0767, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0238, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0560, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0728, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0706, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3911, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2384, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0526, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8997, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2468, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1400, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0257, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0867, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1719, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3443, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 6---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9178082191780822, 'agreement': 0.5812807881773399, 'direct_attack': 0.28888888888888886, 'undercutter_attack': 0.45394736842105265, 'partial': 0.2077922077922078}, 'recall': {'support': 0.9236479321314952, 'agreement': 0.6178010471204188, 'direct_attack': 0.3333333333333333, 'undercutter_attack': 0.44805194805194803, 'partial': 0.14414414414414414}, 'f1': {'support': 0.9207188160676533, 'agreement': 0.5989847715736041, 'direct_attack': 0.3095238095238095, 'undercutter_attack': 0.45098039215686275, 'partial': 0.1702127659574468}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7341688654353562, 'recall': 0.7341688654353562, 'f1': 0.7341688654353562, 'support': None}, 'macro_avg': {'precision': 0.4899434944915143, 'recall': 0.49339568095626796, 'f1': 0.49008411105587524, 'support': None}, 'weighted_avg': {'precision': 0.7277642552958975, 'recall': 0.7341688654353562, 'f1': 0.7303448667876481, 'support': None}}
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0842, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0864, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3823, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1684, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0274, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0752, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0290, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0428, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0414, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0576, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0932, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0849, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0709, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0399, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0623, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2365, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2754, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3410, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0885, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1662, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0575, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0883, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2574, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0508, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2556, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8404, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1482, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0999, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0903, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0444, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8331, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4988, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1465, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2504, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0819, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4727, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2777, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1358, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0900, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0590, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1743, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0670, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0459, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0586, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0769, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0781, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1859, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0506, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0236, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0158, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4502, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0584, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0334, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2672, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1595, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1922, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8697, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0421, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0611, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0617, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1486, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2407, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0802, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0640, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0680, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0222, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3756, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4946, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0954, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0969, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0836, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2796, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4439, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2973, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0964, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0633, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0264, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8395, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0811, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4470, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2758, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0779, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0722, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0643, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3416, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1921, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0418, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3774, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0247, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3644, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7790, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1220, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1509, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2635, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4936, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0392, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0887, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4449, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2498, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9825, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0424, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0438, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0340, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0380, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7958, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0759, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8913, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4955, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0736, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0788, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1760, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5649, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0939, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 7---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9377382465057179, 'agreement': 0.3919308357348703, 'direct_attack': 0.30952380952380953, 'undercutter_attack': 0.4057971014492754, 'partial': 0.18681318681318682}, 'recall': {'support': 0.782608695652174, 'agreement': 0.7120418848167539, 'direct_attack': 0.2222222222222222, 'undercutter_attack': 0.5454545454545454, 'partial': 0.15315315315315314}, 'f1': {'support': 0.8531791907514451, 'agreement': 0.5055762081784386, 'direct_attack': 0.25870646766169153, 'undercutter_attack': 0.46537396121883656, 'partial': 0.1683168316831683}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.6602902374670184, 'recall': 0.6602902374670184, 'f1': 0.6602902374670184, 'support': None}, 'macro_avg': {'precision': 0.44636063600537207, 'recall': 0.4830961002597697, 'f1': 0.450230531898716, 'support': None}, 'weighted_avg': {'precision': 0.711470487568595, 'recall': 0.6602902374670184, 'f1': 0.673965994526151, 'support': None}}
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0975, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1379, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1448, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0363, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1406, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0847, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6511, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0669, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5738, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0582, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0694, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3850, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0641, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0393, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0239, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3868, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0403, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0553, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2830, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0232, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2675, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0745, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0729, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0139, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0852, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2569, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1604, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0534, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0372, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5768, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0285, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0218, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7832, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0685, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0245, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0543, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2980, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0776, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0185, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0341, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0619, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3455, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9622, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5945, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1935, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4599, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3681, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0199, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0843, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0646, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0292, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0637, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0246, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1965, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3748, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1888, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0735, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1522, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0354, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2902, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7297, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0609, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1464, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1376, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2388, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3703, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0219, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0353, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0413, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0803, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0630, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0949, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0197, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2624, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1417, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0546, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0314, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0579, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1311, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0772, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0533, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0338, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4552, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1928, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2631, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6231, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0113, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0426, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0456, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0360, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5436, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0557, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7261, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0540, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2833, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1327, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0572, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4391, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0305, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0323, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 8---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9272116461366181, 'agreement': 0.46621621621621623, 'direct_attack': 0.25301204819277107, 'undercutter_attack': 0.4056603773584906, 'partial': 0.28125}, 'recall': {'support': 0.8780487804878049, 'agreement': 0.7225130890052356, 'direct_attack': 0.1794871794871795, 'undercutter_attack': 0.5584415584415584, 'partial': 0.08108108108108109}, 'f1': {'support': 0.9019607843137255, 'agreement': 0.566735112936345, 'direct_attack': 0.21000000000000002, 'undercutter_attack': 0.46994535519125685, 'partial': 0.1258741258741259}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7137203166226913, 'recall': 0.7137203166226913, 'f1': 0.7137203166226912, 'support': None}, 'macro_avg': {'precision': 0.46667005758081925, 'recall': 0.4839143377005719, 'f1': 0.45490307566309063, 'support': None}, 'weighted_avg': {'precision': 0.7168210668574472, 'recall': 0.7137203166226913, 'f1': 0.7056128224605321, 'support': None}}
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1794, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0134, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7753, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2475, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1362, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1241, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0301, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3342, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1815, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0671, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1435, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0717, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0422, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1499, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2798, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8953, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3871, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1916, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4821, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6749, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3332, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0668, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0324, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1667, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2295, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4000, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0432, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0817, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0478, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5740, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1378, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1784, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0771, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0786, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0495, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9831, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4863, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2344, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0912, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0457, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6386, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2818, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8876, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2425, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0601, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8437, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4442, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0799, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3577, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1792, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1881, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0460, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1229, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3711, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7742, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0995, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3542, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6854, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2263, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0366, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2283, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1262, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4634, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1762, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0647, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4204, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0568, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1658, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2221, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0452, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0481, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1739, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0322, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4713, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7877, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9806, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0195, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3682, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0308, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0585, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0571, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3983, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2807, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0278, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1500, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0925, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2510, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0840, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1919, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8474, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0345, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0174, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0121, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0429, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1686, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9352, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2381, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2210, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0878, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1981, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1434, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0677, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1300, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6937, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1665, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0541, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1775, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1512, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0596, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0317, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0691, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0861, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0966, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8848, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9489, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0592, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3687, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0294, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2328, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0989, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0610, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6412, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0853, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3612, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4284, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1537, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0503, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0494, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0720, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7520, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0213, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2679, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0313, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0692, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0351, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0493, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0463, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0251, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0785, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3683, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0375, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0458, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8154, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1613, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0587, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0639, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0205, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0472, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0656, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1702, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0408, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0152, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0483, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0364, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2361, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0491, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1548, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0731, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0155, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4279, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0480, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0230, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0433, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0315, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0272, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2531, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0761, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0805, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2918, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0896, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0476, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0940, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0088, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0163, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0419, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3527, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1501, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4789, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6291, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1565, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0562, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6800, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0269, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0856, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0242, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9917, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0757, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0524, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6288, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0467, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0369, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0638, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8420, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4914, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7795, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1661, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7904, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0293, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0370, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7907, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2956, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0423, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5924, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0387, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3747, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1496, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1827, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1620, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1549, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3898, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1226, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0469, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0347, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4891, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0310, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2450, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0783, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0536, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3770, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1336, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 9---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.8931841302136317, 'agreement': 0.5138121546961326, 'direct_attack': 0.24571428571428572, 'undercutter_attack': 0.5089285714285714, 'partial': 0.15384615384615385}, 'recall': {'support': 0.9310710498409331, 'agreement': 0.4869109947643979, 'direct_attack': 0.36752136752136755, 'undercutter_attack': 0.37012987012987014, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9117341640706126, 'agreement': 0.5000000000000001, 'direct_attack': 0.2945205479452055, 'undercutter_attack': 0.42857142857142855, 'partial': 0.11363636363636363}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7130606860158312, 'recall': 0.7130606860158312, 'f1': 0.7130606860158312, 'support': None}, 'macro_avg': {'precision': 0.463097059179755, 'recall': 0.44914467446933176, 'f1': 0.44969250084472207, 'support': None}, 'weighted_avg': {'precision': 0.702250165464321, 'recall': 0.7130606860158312, 'f1': 0.7047083490711168, 'support': None}}
Loss: tensor(0.0614, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0588, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0398, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0240, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0191, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0194, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0098, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0479, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5938, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0216, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0179, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0306, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0972, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0723, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0234, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0716, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0356, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0256, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1273, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0605, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0227, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1651, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1737, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0212, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0377, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0755, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0523, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0615, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0159, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5337, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2629, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1477, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1348, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0181, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2744, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5248, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0186, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0857, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1280, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2454, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1519, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0451, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2894, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0357, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1708, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0547, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5371, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0359, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0987, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0177, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3427, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0111, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0724, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4678, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0402, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1521, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0307, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8869, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0201, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0237, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0858, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0165, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0791, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0252, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1905, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1277, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1695, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0538, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0215, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0172, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0979, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0224, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0289, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0302, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0266, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0196, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0282, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0276, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0147, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0461, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0978, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0721, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0318, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0666, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0209, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0514, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0725, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0286, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0397, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0532, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4976, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0698, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3453, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1389, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0243, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1688, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6970, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0110, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0942, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0330, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0730, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0255, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0657, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5244, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0265, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0934, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0130, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0346, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0544, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0148, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0250, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1497, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0156, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0208, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0225, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0696, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3267, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0355, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2184, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0160, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7645, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2440, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0296, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.7948, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0367, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0325, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0368, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0385, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1616, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2923, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0136, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0203, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0125, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0180, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0712, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0594, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0126, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0333, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3841, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0249, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0173, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0598, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0445, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0162, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0182, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0608, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0895, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0304, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1985, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.8003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.5409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0258, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0545, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0573, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0095, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0183, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1228, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1447, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0415, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0880, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.9897, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0153, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0350, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 10---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9107142857142857, 'agreement': 0.54337899543379, 'direct_attack': 0.2740740740740741, 'undercutter_attack': 0.5083798882681564, 'partial': 0.22580645161290322}, 'recall': {'support': 0.9194061505832449, 'agreement': 0.6230366492146597, 'direct_attack': 0.3162393162393162, 'undercutter_attack': 0.5909090909090909, 'partial': 0.06306306306306306}, 'f1': {'support': 0.9150395778364115, 'agreement': 0.5804878048780489, 'direct_attack': 0.2936507936507936, 'undercutter_attack': 0.5465465465465466, 'partial': 0.09859154929577464}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7394459102902374, 'recall': 0.7394459102902374, 'f1': 0.7394459102902375, 'support': None}, 'macro_avg': {'precision': 0.49247073902064187, 'recall': 0.5025308540018749, 'f1': 0.486863254441515, 'support': None}, 'weighted_avg': {'precision': 0.7242814281961876, 'recall': 0.7394459102902374, 'f1': 0.7277206237655577, 'support': None}}
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0287, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0143, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0223, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0135, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0092, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0298, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0206, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0466, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0117, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0401, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0146, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0193, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2149, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0260, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0329, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0161, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.2673, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0275, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0910, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0992, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0108, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0235, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0140, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0321, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0106, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0409, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3971, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0253, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0132, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0984, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.4189, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0233, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0198, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0593, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0145, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0808, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0254, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0175, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0167, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0200, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0664, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0642, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0128, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0636, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0487, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0259, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0689, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0214, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0141, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0103, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0083, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0137, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0142, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.1714, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0104, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0157, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0129, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0133, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0710, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0507, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3396, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0122, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 11---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9140127388535032, 'agreement': 0.5486725663716814, 'direct_attack': 0.2620689655172414, 'undercutter_attack': 0.5029239766081871, 'partial': 0.25}, 'recall': {'support': 0.9130434782608695, 'agreement': 0.6492146596858639, 'direct_attack': 0.3247863247863248, 'undercutter_attack': 0.5584415584415584, 'partial': 0.07207207207207207}, 'f1': {'support': 0.9135278514588859, 'agreement': 0.5947242206235012, 'direct_attack': 0.29007633587786263, 'undercutter_attack': 0.5292307692307692, 'partial': 0.11188811188811189}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7368073878627969, 'recall': 0.7368073878627969, 'f1': 0.7368073878627969, 'support': None}, 'macro_avg': {'precision': 0.4955356494701227, 'recall': 0.5035116186493378, 'f1': 0.48788945781582627, 'support': None}, 'weighted_avg': {'precision': 0.7272907877829966, 'recall': 0.7368073878627969, 'f1': 0.7275126254905323, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0078, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0094, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0085, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0281, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0217, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0131, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0107, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0202, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0349, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0558, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0093, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0079, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0075, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0271, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0190, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0176, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0116, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0187, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1170, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.6741, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0178, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0124, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0127, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0120, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0168, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6509e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0119, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0578, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0071, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0063, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0309, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0069, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0320, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7214e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0099, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0096, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 12---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9121693121693122, 'agreement': 0.5466666666666666, 'direct_attack': 0.2755905511811024, 'undercutter_attack': 0.489247311827957, 'partial': 0.2727272727272727}, 'recall': {'support': 0.9141039236479321, 'agreement': 0.643979057591623, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5909090909090909, 'partial': 0.08108108108108109}, 'f1': {'support': 0.913135593220339, 'agreement': 0.5913461538461537, 'direct_attack': 0.2868852459016394, 'undercutter_attack': 0.5352941176470589, 'partial': 0.125}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7387862796833773, 'recall': 0.7387862796833773, 'f1': 0.7387862796833774, 'support': None}, 'macro_avg': {'precision': 0.49928022291446206, 'recall': 0.5058436904750053, 'f1': 0.4903322221230383, 'support': None}, 'weighted_avg': {'precision': 0.7272096982133353, 'recall': 0.7387862796833773, 'f1': 0.7281727227437559, 'support': None}}
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0625, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.3591, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0164, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0049, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0169, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0166, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0072, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0097, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0118, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0112, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0087, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0150, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0109, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0077, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0058, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0115, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0151, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 13---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9103375527426161, 'agreement': 0.55, 'direct_attack': 0.28225806451612906, 'undercutter_attack': 0.4842105263157895, 'partial': 0.2647058823529412}, 'recall': {'support': 0.9151643690349947, 'agreement': 0.6335078534031413, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5974025974025974, 'partial': 0.08108108108108109}, 'f1': {'support': 0.9127445795875199, 'agreement': 0.5888077858880779, 'direct_attack': 0.29045643153526973, 'undercutter_attack': 0.5348837209302325, 'partial': 0.12413793103448276}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7387862796833773, 'recall': 0.7387862796833773, 'f1': 0.7387862796833774, 'support': None}, 'macro_avg': {'precision': 0.4983024051854952, 'recall': 0.5052602400134227, 'f1': 0.49020608979511654, 'support': None}, 'weighted_avg': {'precision': 0.7259058573736689, 'recall': 0.7387862796833773, 'f1': 0.7277804957212163, 'support': None}}
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0102, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0138, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0192, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0114, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6437e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0091, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0057, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0082, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0068, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0105, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0054, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0081, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0080, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0070, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0123, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0061, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0303, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7636e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 14---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.91129883843717, 'agreement': 0.5520361990950227, 'direct_attack': 0.2777777777777778, 'undercutter_attack': 0.47593582887700536, 'partial': 0.2571428571428571}, 'recall': {'support': 0.9151643690349947, 'agreement': 0.6387434554973822, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.577922077922078, 'partial': 0.08108108108108109}, 'f1': {'support': 0.913227513227513, 'agreement': 0.5922330097087379, 'direct_attack': 0.2880658436213992, 'undercutter_attack': 0.5219941348973607, 'partial': 0.12328767123287672}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.737467018469657, 'recall': 0.737467018469657, 'f1': 0.737467018469657, 'support': None}, 'macro_avg': {'precision': 0.4948383002659666, 'recall': 0.502411256536167, 'f1': 0.4877616345375776, 'support': None}, 'weighted_avg': {'precision': 0.7250202463478342, 'recall': 0.737467018469657, 'f1': 0.7269563204569, 'support': None}}
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0060, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0089, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0144, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0065, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0064, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0100, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0090, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0101, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0066, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0042, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 15---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.91129883843717, 'agreement': 0.5520361990950227, 'direct_attack': 0.2698412698412698, 'undercutter_attack': 0.4731182795698925, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9151643690349947, 'agreement': 0.6387434554973822, 'direct_attack': 0.2905982905982906, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.913227513227513, 'agreement': 0.5922330097087379, 'direct_attack': 0.27983539094650206, 'undercutter_attack': 0.5176470588235295, 'partial': 0.1360544217687075}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7368073878627969, 'recall': 0.7368073878627969, 'f1': 0.7368073878627969, 'support': None}, 'macro_avg': {'precision': 0.49681447294422654, 'recall': 0.5012049553298658, 'f1': 0.48779947889499803, 'support': None}, 'weighted_avg': {'precision': 0.7256323849814813, 'recall': 0.7368073878627969, 'f1': 0.7268142997650426, 'support': None}}
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0086, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0074, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0076, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1341e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0067, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0053, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0043, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0055, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0059, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0048, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0084, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2896e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 16---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9114857744994731, 'agreement': 0.5570776255707762, 'direct_attack': 0.2677165354330709, 'undercutter_attack': 0.4756756756756757, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9172852598091198, 'agreement': 0.6387434554973822, 'direct_attack': 0.2905982905982906, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9143763213530653, 'agreement': 0.5951219512195122, 'direct_attack': 0.27868852459016397, 'undercutter_attack': 0.5191740412979351, 'partial': 0.1360544217687075}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7381266490765171, 'recall': 0.7381266490765171, 'f1': 0.7381266490765173, 'support': None}, 'macro_avg': {'precision': 0.4979466777913547, 'recall': 0.5016291334846908, 'f1': 0.4886830520458769, 'support': None}, 'weighted_avg': {'precision': 0.7264796397559881, 'recall': 0.7381266490765171, 'f1': 0.727959475113539, 'support': None}}
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0050, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3611e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8816e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0037, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8820e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3482e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0073, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1359e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0041, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 17---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9114857744994731, 'agreement': 0.5570776255707762, 'direct_attack': 0.2677165354330709, 'undercutter_attack': 0.4756756756756757, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9172852598091198, 'agreement': 0.6387434554973822, 'direct_attack': 0.2905982905982906, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9143763213530653, 'agreement': 0.5951219512195122, 'direct_attack': 0.27868852459016397, 'undercutter_attack': 0.5191740412979351, 'partial': 0.1360544217687075}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7381266490765171, 'recall': 0.7381266490765171, 'f1': 0.7381266490765173, 'support': None}, 'macro_avg': {'precision': 0.4979466777913547, 'recall': 0.5016291334846908, 'f1': 0.4886830520458769, 'support': None}, 'weighted_avg': {'precision': 0.7264796397559881, 'recall': 0.7381266490765171, 'f1': 0.727959475113539, 'support': None}}
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0039, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0031, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4781e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5650e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6039e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0062, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2655e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0036, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0045, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0056, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0051, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 18---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9114857744994731, 'agreement': 0.555045871559633, 'direct_attack': 0.2677165354330709, 'undercutter_attack': 0.4756756756756757, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9172852598091198, 'agreement': 0.6335078534031413, 'direct_attack': 0.2905982905982906, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9143763213530653, 'agreement': 0.5916870415647921, 'direct_attack': 0.27868852459016397, 'undercutter_attack': 0.5191740412979351, 'partial': 0.13513513513513514}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.737467018469657, 'recall': 0.737467018469657, 'f1': 0.737467018469657, 'support': None}, 'macro_avg': {'precision': 0.49603882548762457, 'recall': 0.5005820130658426, 'f1': 0.48781221278821835, 'support': None}, 'weighted_avg': {'precision': 0.7256739680215148, 'recall': 0.737467018469657, 'f1': 0.727459403503791, 'support': None}}
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0602e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4296e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0044, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0052, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6579e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0047, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7456e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4178e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0033, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 19---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9113924050632911, 'agreement': 0.555045871559633, 'direct_attack': 0.2734375, 'undercutter_attack': 0.4756756756756757, 'partial': 0.2702702702702703}, 'recall': {'support': 0.9162248144220573, 'agreement': 0.6335078534031413, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.913802221047065, 'agreement': 0.5916870415647921, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.5191740412979351, 'partial': 0.13513513513513514}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.737467018469657, 'recall': 0.737467018469657, 'f1': 0.737467018469657, 'support': None}, 'macro_avg': {'precision': 0.497164344513774, 'recall': 0.5020793256978319, 'f1': 0.4891025449518425, 'support': None}, 'weighted_avg': {'precision': 0.7260574149054272, 'recall': 0.737467018469657, 'f1': 0.7276445205637936, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2676e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7665e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0029, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0397e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3669e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0032, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2207e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0038, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0966e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9392e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5489e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5819e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1035e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0046, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9120e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3751e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8059e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 20---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9104320337197049, 'agreement': 0.5529953917050692, 'direct_attack': 0.2734375, 'undercutter_attack': 0.4731182795698925, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9162248144220573, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9133192389006343, 'agreement': 0.5882352941176471, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.5176470588235295, 'partial': 0.1360544217687075}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7368073878627969, 'recall': 0.7368073878627969, 'f1': 0.7368073878627969, 'support': None}, 'macro_avg': {'precision': 0.4975521965544889, 'recall': 0.5010322052789837, 'f1': 0.48819405986496084, 'support': None}, 'weighted_avg': {'precision': 0.7254915986150704, 'recall': 0.7368073878627969, 'f1': 0.7268214002397693, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2886e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2310e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6182e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3762e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0027, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0035, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3752e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0823e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0040, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5375e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0581e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0026, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2731e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7127e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5304e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0021, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 21---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9104320337197049, 'agreement': 0.5529953917050692, 'direct_attack': 0.2734375, 'undercutter_attack': 0.4731182795698925, 'partial': 0.2777777777777778}, 'recall': {'support': 0.9162248144220573, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9133192389006343, 'agreement': 0.5882352941176471, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.5176470588235295, 'partial': 0.1360544217687075}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7368073878627969, 'recall': 0.7368073878627969, 'f1': 0.7368073878627969, 'support': None}, 'macro_avg': {'precision': 0.4975521965544889, 'recall': 0.5010322052789837, 'f1': 0.48819405986496084, 'support': None}, 'weighted_avg': {'precision': 0.7254915986150704, 'recall': 0.7368073878627969, 'f1': 0.7268214002397693, 'support': None}}
Loss: tensor(4.7609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0034, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4123e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8098e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0028, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2068e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4118e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5813e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4875e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6872e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1117e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 22---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9104320337197049, 'agreement': 0.5529953917050692, 'direct_attack': 0.2734375, 'undercutter_attack': 0.47593582887700536, 'partial': 0.2857142857142857}, 'recall': {'support': 0.9162248144220573, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.577922077922078, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9133192389006343, 'agreement': 0.5882352941176471, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.5219941348973607, 'partial': 0.13698630136986303}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.737467018469657, 'recall': 0.737467018469657, 'f1': 0.737467018469657, 'support': None}, 'macro_avg': {'precision': 0.499703008003213, 'recall': 0.502330906577685, 'f1': 0.4892498509999582, 'support': None}, 'weighted_avg': {'precision': 0.7263589171996666, 'recall': 0.737467018469657, 'f1': 0.7273312210518394, 'support': None}}
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0024, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6850e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2149e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9906e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3163e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2691e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7458e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8876e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0030, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0022, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1459e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0572e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2398e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2423e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0680e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 23---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9104320337197049, 'agreement': 0.5529953917050692, 'direct_attack': 0.2734375, 'undercutter_attack': 0.4734042553191489, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9162248144220573, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.577922077922078, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9133192389006343, 'agreement': 0.5882352941176471, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.52046783625731, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.737467018469657, 'recall': 0.737467018469657, 'f1': 0.737467018469657, 'support': None}, 'macro_avg': {'precision': 0.5008773655605493, 'recall': 0.502330906577685, 'f1': 0.4891335378945271, 'support': None}, 'weighted_avg': {'precision': 0.7267170377678288, 'recall': 0.737467018469657, 'f1': 0.7272453472952192, 'support': None}}
Loss: tensor(9.1939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6878e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1134e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0888e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4697e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7577e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4910e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3965e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0599e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3336e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4019e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4305e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7332e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6061e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7583e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1371e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6848e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5653e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2335e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0025, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4515e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2085e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5092e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0027e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 24---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9104320337197049, 'agreement': 0.5529953917050692, 'direct_attack': 0.2734375, 'undercutter_attack': 0.4734042553191489, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9162248144220573, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.577922077922078, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9133192389006343, 'agreement': 0.5882352941176471, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.52046783625731, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.737467018469657, 'recall': 0.737467018469657, 'f1': 0.737467018469657, 'support': None}, 'macro_avg': {'precision': 0.5008773655605493, 'recall': 0.502330906577685, 'f1': 0.4891335378945271, 'support': None}, 'weighted_avg': {'precision': 0.7267170377678288, 'recall': 0.737467018469657, 'f1': 0.7272453472952192, 'support': None}}
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0023, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0633e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7250e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3842e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4811e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5109e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0088e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2206e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1258e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3912e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0936e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0018, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5642e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5666e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3885e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6463e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3056e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9249e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9279e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5486e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1844e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3366e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8787e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9560e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6861e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1008e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1334e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8071e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7911e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3747e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5631e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8379e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3841e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2527e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0308e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0771e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2951e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2497e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3729e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4511e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 25---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9103375527426161, 'agreement': 0.5504587155963303, 'direct_attack': 0.2734375, 'undercutter_attack': 0.4734042553191489, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9151643690349947, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.577922077922078, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9127445795875199, 'agreement': 0.58679706601467, 'direct_attack': 0.2857142857142857, 'undercutter_attack': 0.52046783625731, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7368073878627969, 'recall': 0.7368073878627969, 'f1': 0.7368073878627969, 'support': None}, 'macro_avg': {'precision': 0.5003511341433837, 'recall': 0.5021188175002724, 'f1': 0.48873096041130876, 'support': None}, 'weighted_avg': {'precision': 0.7263386731912035, 'recall': 0.7368073878627969, 'f1': 0.7267066894456574, 'support': None}}
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8754e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6766e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0020, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7674e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5857e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7531e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2238e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5934e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3699e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0151e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9284e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7908e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8484e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6242e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4183e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3339e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7612e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0298e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0017, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8492e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2116e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0542e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8903e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6784e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2097e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8280e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3060e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0635e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7743e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9179e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2663e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7311e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7129e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3790e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5681e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8426e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0393e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7105e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9913e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1753e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1450e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8591e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4773e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7864e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2682e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5084e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5156e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4115e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3033e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0148e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0922e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8795e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5767e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2451e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 26---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9103375527426161, 'agreement': 0.5504587155963303, 'direct_attack': 0.2755905511811024, 'undercutter_attack': 0.4708994708994709, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9151643690349947, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.577922077922078, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9127445795875199, 'agreement': 0.58679706601467, 'direct_attack': 0.2868852459016394, 'undercutter_attack': 0.5189504373177842, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7368073878627969, 'recall': 0.7368073878627969, 'f1': 0.7368073878627969, 'support': None}, 'macro_avg': {'precision': 0.5002807874956685, 'recall': 0.5021188175002724, 'f1': 0.4886616726608744, 'support': None}, 'weighted_avg': {'precision': 0.7262503949508067, 'recall': 0.7368073878627969, 'f1': 0.7266429182749671, 'support': None}}
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3760e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2390e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6248e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1755e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0015, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0270e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9005e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4746e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9218e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0247e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7712e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1539e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3873e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9287e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8937e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3839e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7890e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5252e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8110e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9020e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2588e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6978e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7698e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1761e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1701e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2086e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4728e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0831e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0048e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2921e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3428e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8164e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0983e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1338e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9724e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5246e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1744e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9997e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5637e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2828e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2609e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5064e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0012, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0019, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3764e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2830e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2661e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5230e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0981e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0301e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0693e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0157e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7002e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8122e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6554e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0598e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2312e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6887e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6502e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3938e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4011e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9969e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8431e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 27---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9103375527426161, 'agreement': 0.5504587155963303, 'direct_attack': 0.2713178294573643, 'undercutter_attack': 0.47058823529411764, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9151643690349947, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9127445795875199, 'agreement': 0.58679706601467, 'direct_attack': 0.2845528455284553, 'undercutter_attack': 0.5161290322580646, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7361477572559367, 'recall': 0.7361477572559367, 'f1': 0.7361477572559366, 'support': None}, 'macro_avg': {'precision': 0.4993639960298504, 'recall': 0.5008201162015712, 'f1': 0.4876309115742936, 'support': None}, 'weighted_avg': {'precision': 0.7258890237602382, 'recall': 0.7361477572559367, 'f1': 0.7261763040118671, 'support': None}}
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0822e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1481e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9276e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1052e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6640e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2702e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5525e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4079e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2445e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2710e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9128e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6424e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3420e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7717e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4201e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0016, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3902e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2233e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7632e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0402e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8825e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7195e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2236e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0126e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1172e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9213e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7522e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4025e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0791e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5580e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7549e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5114e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9417e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8522e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1770e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2051e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7726e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2920e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2576e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9012e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1021e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1013e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0443e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1968e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4815e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0011, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3524e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2604e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7641e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6295e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2972e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5161e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9681e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6069e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9617e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6598e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8468e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5411e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7704e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1718e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8171e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1454e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4009e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1763e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0961e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5386e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5957e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7440e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4798e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7914e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1496e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2277e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4356e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5740e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2315e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7946e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0789e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4867e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8103e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7480e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4664e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5040e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4931e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2709e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1851e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7349e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1457e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3291e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3357e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6044e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3327e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7220e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5705e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9119e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3188e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9530e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0368e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5948e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3865e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6769e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 28---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9102428722280888, 'agreement': 0.547945205479452, 'direct_attack': 0.2713178294573643, 'undercutter_attack': 0.47058823529411764, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9141039236479321, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9121693121693123, 'agreement': 0.5853658536585366, 'direct_attack': 0.2845528455284553, 'undercutter_attack': 0.5161290322580646, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7354881266490765, 'recall': 0.7354881266490765, 'f1': 0.7354881266490764, 'support': None}, 'macro_avg': {'precision': 0.49884235790356923, 'recall': 0.5006080271241586, 'f1': 0.48722961561942546, 'support': None}, 'weighted_avg': {'precision': 0.7255134537354869, 'recall': 0.7354881266490765, 'f1': 0.7256381518117411, 'support': None}}
Loss: tensor(4.2540e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1723e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6342e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4253e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3466e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4382e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4184e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6495e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1827e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8998e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6199e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8112e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6281e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1467e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6996e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9919e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3932e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6121e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6000e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8181e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3058e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9093e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7949e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5273e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3062e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9045e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6730e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5084e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3449e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6494e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7757e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6372e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2373e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5321e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7493e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8192e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2460e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5503e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7306e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8297e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4212e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1074e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6037e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0014, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2293e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6169e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9067e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3235e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6303e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1649e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7506e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3607e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4189e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5738e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1589e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8907e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5343e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5501e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1606e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3404e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8601e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4871e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3186e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5962e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9340e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3578e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1477e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1429e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9360e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5689e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5209e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3868e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9573e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4781e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8514e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4538e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4041e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3295e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7592e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3889e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4500e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6785e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8470e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3992e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3881e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3570e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3859e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6590e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8883e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5291e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2933e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9984e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3275e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2406e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3807e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5659e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6023e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9300e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9625e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4809e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9160e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4289e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9893e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7448e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7223e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9832e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0523e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2626e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4378e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6102e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4034e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6210e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5869e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4244e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3401e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8104e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7138e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4898e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0290e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3647e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8003e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8542e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6350e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3016e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7955e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9952e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7313e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0237e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7742e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2006e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6024e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6146e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9422e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3678e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0861e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1516e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2194e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9824e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5147e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7519e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2838e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5847e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3776e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0826e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1732e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2367e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4789e-06, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 29---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9102428722280888, 'agreement': 0.547945205479452, 'direct_attack': 0.27692307692307694, 'undercutter_attack': 0.4731182795698925, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9141039236479321, 'agreement': 0.6282722513089005, 'direct_attack': 0.3076923076923077, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9121693121693123, 'agreement': 0.5853658536585366, 'direct_attack': 0.29149797570850206, 'undercutter_attack': 0.5176470588235295, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7361477572559367, 'recall': 0.7361477572559367, 'f1': 0.7361477572559366, 'support': None}, 'macro_avg': {'precision': 0.5004694162518667, 'recall': 0.5023174288335603, 'f1': 0.48892224696852776, 'support': None}, 'weighted_avg': {'precision': 0.7262030584663298, 'recall': 0.7361477572559367, 'f1': 0.7263283604675109, 'support': None}}
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9916e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2964e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7786e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1395e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4782e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2964e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0688e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2000e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0901e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3152e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5351e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7038e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9615e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4517e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0022e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3333e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1316e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7792e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2761e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9435e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6472e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2520e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0319e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7999e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9036e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9590e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1564e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8550e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3950e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5533e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7834e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7825e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8211e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9974e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8103e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8409e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1450e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4749e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3436e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8111e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2354e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2419e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4183e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0251e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6963e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0765e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5376e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7007e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2048e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9944e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1165e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4617e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4365e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4628e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2644e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9881e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5836e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9329e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0239e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0991e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4470e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5563e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2510e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3864e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5055e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0892e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1027e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1441e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5314e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6311e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9331e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5725e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6566e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5603e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1309e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8328e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9845e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4274e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3042e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5896e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8544e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4806e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6229e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0802e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1982e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8709e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3562e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2512e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6203e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9090e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5031e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4132e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6345e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1269e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4261e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6553e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0438e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8032e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5389e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3425e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5697e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8706e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4396e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1993e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7139e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4917e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2719e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2176e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0837e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4623e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0196e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9205e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8197e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8400e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3958e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6221e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5600e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4504e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7219e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6145e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2536e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0880e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0840e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5977e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0231e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1884e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8734e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3141e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5758e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5391e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9745e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1247e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4222e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2596e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8756e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0534e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6865e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2488e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1104e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6100e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9772e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7877e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0013, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.8406e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6012e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8733e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0446e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6403e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1267e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0658e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9942e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8271e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5125e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0065e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2172e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5089e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6645e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1063e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5796e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3485e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8228e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9054e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6660e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0174e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7388e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0355e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8793e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0004e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0559e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4662e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3595e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1175e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5646e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5282e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4167e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9029e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5010e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0263e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4170e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7107e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4685e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3667e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1081e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0195e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4486e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6108e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2056e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8703e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5442e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3953e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8725e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7318e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1173e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6849e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0256e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0204e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3193e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6905e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5330e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1586e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8622e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6191e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8773e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6696e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0394e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5879e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3638e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9285e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7073e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1087e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1264e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0083e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0862e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5924e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0166e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9478e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2351e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2407e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3344e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4689e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6736e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3361e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5092e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8621e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1558e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1288e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0608e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9584e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5464e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1243e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1082e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4927e-05, device='cuda:0', grad_fn=<DivBackward0>)
			------------EPOCH 30---------------
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9102428722280888, 'agreement': 0.547945205479452, 'direct_attack': 0.2755905511811024, 'undercutter_attack': 0.4656084656084656, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9141039236479321, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9121693121693123, 'agreement': 0.5853658536585366, 'direct_attack': 0.2868852459016394, 'undercutter_attack': 0.5131195335276968, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7354881266490765, 'recall': 0.7354881266490765, 'f1': 0.7354881266490764, 'support': None}, 'macro_avg': {'precision': 0.4987009483111864, 'recall': 0.5006080271241586, 'f1': 0.4870941959479887, 'support': None}, 'weighted_avg': {'precision': 0.7253373481352804, 'recall': 0.7354881266490765, 'f1': 0.7255124447135788, 'support': None}}
Loss: tensor(1.7259e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0944e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4808e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4534e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0768e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0521e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9714e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9348e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3241e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3462e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4616e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2278e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9945e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4153e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4555e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5979e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3570e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7895e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9240e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2727e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0891e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8167e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6894e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7733e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3610e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5927e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9106e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3353e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6614e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4175e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5505e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7078e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9928e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3254e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9633e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9185e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3769e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5619e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4971e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2679e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6131e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9975e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1597e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5453e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2094e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1294e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4671e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1960e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1476e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8099e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5018e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7814e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4461e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0818e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7672e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1990e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4858e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5873e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3283e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6708e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5546e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0075e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5260e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0803e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8673e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7777e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0430e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5224e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4614e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3133e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1139e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.0255e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4473e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7091e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1364e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1952e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3799e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8414e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2904e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4215e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1190e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9136e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7020e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4561e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0317e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5713e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1498e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9852e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2657e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2447e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6017e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1593e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8986e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4797e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2427e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9797e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9339e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1217e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9434e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6805e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3928e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1753e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6513e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4162e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0286e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9651e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5866e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6575e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.1769e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4715e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2043e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5987e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6543e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1587e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7471e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7801e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9028e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2158e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0870e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9854e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7432e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6177e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5984e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8422e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4302e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6926e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2569e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3995e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8245e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3030e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0853e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9283e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0053e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4191e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9571e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5582e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8717e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7380e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9268e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9219e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9386e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5897e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0226e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0399e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6287e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0325e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1800e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2956e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5941e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0788e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3124e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1198e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6320e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1829e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3202e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3578e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8137e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9863e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5255e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8654e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9935e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.6648e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1370e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.1266e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1095e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1337e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4629e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2618e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.9239e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7077e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4821e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7410e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1299e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1748e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0810e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0006, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2367e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.1939e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4323e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0113e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9967e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0066e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8817e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3413e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7475e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2925e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1057e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5720e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1156e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3954e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6956e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4707e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6225e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4140e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9499e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8833e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5568e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6692e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9315e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0014e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5668e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4479e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0537e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0535e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6405e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2860e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6387e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.3257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1385e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5096e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1187e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.3722e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4050e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.6973e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0307e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7856e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.6016e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2490e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3383e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.9613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2331e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.5076e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4630e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9143e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5684e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9408e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5200e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8412e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.5362e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7985e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2216e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3080e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2263e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.2661e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0639e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.9135e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0474e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6272e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.7909e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1778e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9923e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1613e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7846e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0363e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.5049e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4292e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.4418e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.4574e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4322e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6518e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1783e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4478e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1506e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.0452e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.8694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0507e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0900e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2739e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.9920e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.7585e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.9323e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1346e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5976e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5643e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.3690e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1556e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5378e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7947e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6070e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.5150e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0529e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2142e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0416e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7047e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1686e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.3541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.8780e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1415e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4547e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2374e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.4439e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1528e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0487e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9988e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.8759e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0007, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0675e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.7203e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4494e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.1943e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.2980e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7211e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0234e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6154e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1930e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7804e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6900e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9015e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.1551e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.9469e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7168e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1421e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2433e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.5837e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6774e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.9028e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0567e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0779e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.4695e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.1801e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9257e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.0262e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.9620e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2155e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.3208e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7652e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0347e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.1455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.6812e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.7735e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.2130e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.2001e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8915e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.4737e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1111e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.9694e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2711e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4369e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0625e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2384e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.3491e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4352e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.4508e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3565e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2959e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.4775e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3716e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.4980e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2414e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.5882e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0009, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0303e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6541e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1324e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4594e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.2557e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.5624e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8843e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0010, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8159e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2994e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2144e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.8552e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2899e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0741e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.0046e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2275e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.1670e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.5455e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.0265e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5076e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.7444e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.4231e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.7483e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3656e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.5377e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0008, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0532e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(8.8627e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6835e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.2700e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2687e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.3989e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.0677e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.2548e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.0970e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.7227e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0005, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.2653e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(3.6545e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.2929e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.3232e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0003, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.8918e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0004, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(6.6605e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.6026e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7683e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.6465e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0002, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.6381e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(9.3794e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(0.0001, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(7.7514e-06, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(4.8101e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.1112e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(1.7562e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(5.8750e-05, device='cuda:0', grad_fn=<DivBackward0>)
Loss: tensor(2.0855e-05, device='cuda:0', grad_fn=<DivBackward0>)
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
Evaluating
				 {'precision': {'support': 0.9102428722280888, 'agreement': 0.547945205479452, 'direct_attack': 0.2755905511811024, 'undercutter_attack': 0.4656084656084656, 'partial': 0.29411764705882354}, 'recall': {'support': 0.9141039236479321, 'agreement': 0.6282722513089005, 'direct_attack': 0.29914529914529914, 'undercutter_attack': 0.5714285714285714, 'partial': 0.09009009009009009}, 'f1': {'support': 0.9121693121693123, 'agreement': 0.5853658536585366, 'direct_attack': 0.2868852459016394, 'undercutter_attack': 0.5131195335276968, 'partial': 0.13793103448275862}, 'support': {'support': 943, 'agreement': 191, 'direct_attack': 117, 'undercutter_attack': 154, 'partial': 111}, 'micro_avg': {'precision': 0.7354881266490765, 'recall': 0.7354881266490765, 'f1': 0.7354881266490764, 'support': None}, 'macro_avg': {'precision': 0.4987009483111864, 'recall': 0.5006080271241586, 'f1': 0.4870941959479887, 'support': None}, 'weighted_avg': {'precision': 0.7253373481352804, 'recall': 0.7354881266490765, 'f1': 0.7255124447135788, 'support': None}}
